November 2025 arXiv papers — page 113

Zhiyao Ma, In Gim, Lin Zhong

We present Cacheback Decoding, a training-free and model-agnostic speculative decoding method that exploits the locality in language to accelerate Large Language Model (LLM) inference. Cacheback leverages only Least Recently Used (LRU) cache tables of token n-grams to generate draft sequences. Cacheback achieves state-of-the-art performance among comparable

Yikun Li, Matteo Grella, Daniel Nahmias, Gal Engelberg

In recent years, Infrastructure as Code (IaC) has emerged as a critical approach for managing and provisioning IT infrastructure through code and automation. IaC enables organizations to create scalable and consistent environments, effectively managing servers and development settings. However, the growing complexity of cloud infrastructures has led to an in

Weiqi Meng, Hongyi Li, Bai Cui

In the day-ahead energy market, the offering strategy of distributed energy resource (DER) aggregators must be submitted before the uncertainty realization in the form of price-quantity pairs. This work addresses the day-ahead offering problem through a two-stage adaptive robust stochastic optimization model, wherein the first-stage price-quantity pairs and

Sanjar Atamuradov

Meta-learning algorithms enable rapid adaptation to new tasks with minimal data, a critical capability for real-world robotic systems. This paper evaluates Model-Agnostic Meta-Learning (MAML) combined with Trust Region Policy Optimization (TRPO) on the MetaWorld ML10 benchmark, a challenging suite of ten diverse robotic manipulation tasks. We implement and a

Ansh Makwe, Akansh Agrawal, Prateek Jain, Akshan Agrawal

Medical image analysis for complex tasks such as severity grading and disease subtype classification poses significant challenges due to intricate and similar visual patterns among classes, scarcity of labeled data, and variability in expert interpretations. Despite the usefulness of existing attention-based models in capturing complex visual patterns for me

Logan Mann, Nayan Saxena, Sarah Tandon, Chenhao Sun

Negation instructions such as 'do not mention $X$' can paradoxically increase the accessibility of $X$ in human thought, a phenomenon known as ironic rebound. Large language models (LLMs) face the same challenge: suppressing a concept requires internally activating it, which may prime rebound instead of avoidance. We investigated this tension with two experi

Nicholas Gunter, Heiko Kabutz, Kaushik Jayaram

Multilayer piezoelectric polyvinylidene fluoride (PVDF) actuators are a promising approach to enhance performance of soft microrobotic systems. In this work, we develop and characterize multilayer PVDF actuators with parallel voltage distribution across each layer, bridging a unique design space between brittle high-force PZT stacks and compliant but lower-b

Jonas Stein, Maximilian Zorn, Leo Sünkel, Thomas Gabor

Quantum optimization allows for up to exponential quantum speedups for specific, possibly industrially relevant problems. As the key algorithm in this field, we motivate and discuss the Quantum Approximate Optimization Algorithm (QAOA), which can be understood as a slightly generalized version of Quantum Annealing for gate-based quantum computers. We delve i

Chenrui Ma, Xi Xiao, Tianyang Wang, Xiao Wang

Flow Matching has limited ability in achieving one-step generation due to its reliance on learned curved trajectories. Previous studies have attempted to address this limitation by either modifying the coupling distribution to prevent interpolant intersections or introducing consistency and mean-velocity modeling to promote straight trajectory learning. Howe

Samuel Wairimu, Leonardo Horn Iwaya

The rise of Artificial Intelligence (AI) has impacted the development of mobile health (mHealth) apps, most notably with the advent of AI-based chatbots used as ubiquitous ``companions'' for various services, from fitness to mental health assistants. While these mHealth chatbots offer clear benefits, such as personalized health information and predictive dia

Yanxin Peng, Qingping Li, Baodong Wu, Shigang Li

As large language models (LLMs) continue to grow in size and complexity, efficient checkpoint saving\&loading has become crucial for managing storage, memory usage, and fault tolerance in LLM training. The current works do not comprehensively take into account the optimization of these several aspects. This paper proposes a novel checkpoint sparsification an

Yinxiang Wu, Neil M. Davies, Ting Ye

Multivariable Mendelian Randomization (MVMR) estimates the direct causal effects of multiple risk factors on an outcome using genetic variants as instruments. The growing availability of summary-level genetic data has created opportunities to apply MVMR in high-dimensional settings with many strongly correlated candidate risk factors. However, existing metho

Mahsa Mirzargar, Sezer Sorgun, Mohammad Javad Nadjafi Arani

Power-type graphs, such as the power graph, the directed power graph, the enhanced power graph and the difference graph, encode significant information about the internal structure of a finite group. Despite substantial investigation in recent years, the precise relationship between these graphs and the subgroup lattice of the underlying group has remained o

Fan Li, Arun Iyengar, Lanyu Xu

In the field of medical imaging, AI-assisted techniques such as object detection, segmentation, and classification are widely employed to alleviate the workload of physicians and doctors. However, single-task models are predominantly used, overlooking the shared information across tasks. This oversight leads to inefficiencies in real-life applications. In th

Davis Robinson, Kyle A. Hamer, Chelsea Kincaid, Michael Chini

We investigate, by solving the time-dependent Schr\"{o}dinger equation in the single-active-electron approximation in helium, a two-color scheme for tabletop high-order harmonic generation (HHG) that combines a mid-infrared (MIR) driving field with an ultrashort ultraviolet (UV) pulse that could be generated via resonant dispersive wave emission in gas-fille

Yiqing Shen, Chenxiao Fan, Chenjia Li, Mathias Unberath

The goal of text-to-video retrieval is to search large databases for relevant videos based on text queries. Existing methods have progressed to handling explicit queries where the visual content of interest is described explicitly; however, they fail with implicit queries where identifying videos relevant to the query requires reasoning. We introduce reasoni

Chamuditha Jayanga Galappaththige, Jason Lai, Lloyd Windrim, Donald Dansereau

Online Scene Change Detection (SCD) is an extremely challenging problem that requires an agent to detect relevant changes on the fly while observing the scene from unconstrained viewpoints. Existing online SCD methods are significantly less accurate than offline approaches. We present the first online SCD approach that is pose-agnostic, label-free, and ensur

Mohamed Amine Kada Zair

This study explores the relationship between cultural awareness, stereotypes, and communication skills among Algerian participants working or studying in multicultural environments. A quantitative questionnaire was administered to 40 respondents to evaluate their levels of cultural awareness, the presence of stereotypical thinking, and the effectiveness of t

Yiqing Shen, Mathias Unberath

Reasoning segmentation enables open-set object segmentation via implicit text queries, therefore serving as a foundation for embodied agents that should operate autonomously in real-world environments. However, existing methods for reasoning segmentation require multimodal large language models with billions of parameters that exceed the computational capabi

Natalia A. Santos, Marlon Jeske, Antonio A. Chaves

This paper addresses the Quadratic Multiple Constraints Variable-Sized Bin Packing Problem (QMC-VSBPP), a challenging combinatorial optimization problem that generalizes the classical bin packing problem by incorporating multiple capacity dimensions, heterogeneous bin types, and quadratic interaction costs between items. We propose two complementary methods

Eren Erberk Erkul

We identify the relativistic-fluid counterpart of the Unruh effect, in which a comoving probe measures a Thermodynamic Unruh temperature. Frame changes in first-order hydrodynamics are recast as a local, time-dependent hyperbolic rotation in a Rindler-style state space where the instantaneous map between frames is the Thermodynamic Boost and its proper-time

Yiqing Shen, Mathias Unberath

Visual reasoning may require models to interpret images and videos and respond to implicit text queries across diverse output formats, from pixel-level segmentation masks to natural language descriptions. Existing approaches rely on supervised fine-tuning with task-specific architectures. For example, reasoning segmentation, grounding, summarization, and vis

Matteo Battisti

The POEMMA-Balloon with Radio (PBR) is a NASA mission designed to study Ultra-High-Energy Cosmic Rays and Very-High-Energy Neutrinos from a balloon platform. Serving as a precursor to the planned POEMMA satellite mission, PBR will be launched aboard a NASA Super Pressure Balloon for a flight at of 33 km altitude in Spring 2027 from Wanaka, New Zealand. The u

Michael Yang, Shijian Deng, William T. Doan, Kai Wang

Conventional, classification-based AI-generated image detection methods cannot explain why an image is considered real or AI-generated in a way a human expert would, which reduces the trustworthiness and persuasiveness of these detection tools for real-world applications. Leveraging Multimodal Large Language Models (MLLMs) has recently become a trending solu

D. Farias, C. Gall, V. A. Villar, K. Auchettl

Type Ibn supernovae (SNe) are characterized by narrow helium (He I) lines from photons produced by the unshocked circumstellar material (CSM). About 80 SNe Ibn have been discovered to date, and only a handful have extensive observational records. Thus, many open questions regarding the progenitor system and the origin of the CSM remain. Here we investigate p

Leroy D'Souza, Akash Karthikeyan, Yash Vardhan Pant, Sebastian Fischmeister

Hybrid dynamical systems result from the interaction of continuous-variable dynamics with discrete events and encompass various systems such as legged robots, vehicles and aircrafts. Challenges arise when the system's modes are characterized by unobservable (latent) parameters and the events that cause system dynamics to switch between different modes are al

Z. Asmaee, N. Hajirasouliha, K. Azizi

The tensor form factors of the $\Delta^+$ baryon are defined through the matrix element of the tensor current and describe its internal structure and spin distribution. We present the full Lorentz decomposition for the $\Delta^+ \rightarrow \Delta^+$ tensor current matrix element, including all independent structures consistent with Lorentz covariance, the R

Yifan Zhu, Sammie Katt, Samuel Kaski

Despite the explosive growth of AI and the technologies built upon it, predicting and inferring the sub-optimal behavior of users or human collaborators remains a critical challenge. In many cases, such behaviors are not a result of irrationality, but rather a rational decision made given inherent cognitive bounds and biased beliefs about the world. In this

Lubomíra Dvořáková, Edita Pelantová

The factor complexity ${\mathcal C}_{\mathbf u}$ of a sequence ${\mathbf u} = u_0u_1u_2 \cdots$ over a finite alphabet counts the number of factors of length $n$ occurring in $\mathbf u$, i.e., ${\mathcal C}_{\mathbf u}(n) = \#{\mathcal L}_n(\mathbf u)$, where ${\mathcal L}_n({\mathbf u)}= \{u_iu_{i+1}\cdots u_{i+n-1}: i \in \mathbb N\}$. Two factors of ${\m

Rana Faryad Ali, Guillermo Aguilar

Lithium niobate is one of the most sought-after materials for nanophotonic devices, including frequency converters, modulators, and quantum light sources. Integration of lithium niobate into optical devices, however, is hampered by significant top-down fabrication challenges due to its exceptional chemical resistance. Scalable fabrication methods that preser

Rasmit Devkota, John H. Wise

QuaRT is a Python library for quantum simulation of radiative transfer in astrophysical and cosmological problems. It features a novel angular redistribution methodology for lattice Boltzmann methods which improves the isotropy of simulations of objects such as stars in non-scattering media.

Sonu Kumar, Gaël Bastien, Ross H. Colman, Maxim Savinov

We report broadband dielectric spectra of the non-Kramers hexaaluminate PrMgAl\textsubscript{11}O\textsubscript{19}, revealing a pronounced interplay between permittivity and magnetization at cryogenic temperatures. The zero-field dielectric response follows a Barrett-type quantum-paraelectric form, while a broad dielectric anomaly near \SI{5}{K} shows a com

Tatsunari Watanabe, Ma Luo

Over any field of characteristic $0$, we prove that the homotopy exact sequence of algebraic fundamental groups for the universal curve with unordered marked points does not split. The same nonsplitting holds for the universal hyperelliptic curve. Our approach extends Chen's topological result to the profinite setting and relies on the use of relative and co

Yuan-Sen Ting, Alberto Accomazzi, Tirthankar Ghosal, Tuan Dung Nguyen

We present a dataset of 408,590 astrophysics papers from arXiv (astro-ph), spanning 1992 through July 2025. Each paper has been processed through a multi-stage pipeline to produce: (1) structured summaries organized into six semantic sections (Background, Motivation, Methodology, Results, Interpretation, Implication), and (2) concept extraction yielding 9,99

Adnan Naimy, Abdallah Slaoui, Abderrahim Lakhfif, Rachid Ahl Laamara

In this work, we introduce an experimentally viable scheme to enhance the simultaneous estimation precision of the couplings $G_{mc}$ and $G_{mb}$, with a particular focus on the performance of heterodyne detection. By comparing simultaneous and individual estimation strategies, we demonstrate that the simultaneous approach offers a notable advantage in our

Bahareh Golchin, Banafsheh Rekabdar

Detecting anomalies in multivariate time series is essential for monitoring complex industrial systems, where high dimensionality, limited labeled data, and subtle dependencies between sensors cause significant challenges. This paper presents a deep reinforcement learning framework that combines a Variational Autoencoder (VAE), an LSTM-based Deep Q-Network (

Armand Kanga, Etienne Pardoux

We consider a spatial SIR epidemic model where the infectivity of infected individuals depends upon their age of infection, and infections are non local. The domain is an unbounded subset of $\R^d$,and the individuals do not move. We extend our earlier result in \cite{AK-EP}, where the domain was bounded, and prove a law of large numbers as the size of the p

Divya Kiran Kadiyala, Alexandros Daglis

The continual increase of cores on server-grade CPUs raises demands on memory systems, which are constrained by limited off-chip pin and data transfer rate scalability. As a result, high-end processors typically feature lower memory bandwidth per core, at the detriment of memory-intensive workloads. We propose alleviating this challenge by improving the util

Mostafa Nozari, Israel Leyva-Mayorga, Fabio Saggese, Gilberto Berardinelli

This paper tackles the sensing-communication trade-off in integrated sensing and communication (ISAC)-empowered subnetworks for mono-static target localization. We propose a low-complexity iterative node selection algorithm that exploits the spatial diversity of subnetwork deployments and dynamically refines the set of sensing subnetworks to maximize localiz

Zhisheng Zheng, Puyuan Peng, Anuj Diwan, Cong Phuoc Huynh

We introduce VoiceCraft-X, an autoregressive neural codec language model which unifies multilingual speech editing and zero-shot Text-to-Speech (TTS) synthesis across 11 languages: English, Mandarin, Korean, Japanese, Spanish, French, German, Dutch, Italian, Portuguese, and Polish. VoiceCraft-X utilizes the Qwen3 large language model for phoneme-free cross-l

Asmit Bandyopadhyay, Anindita Das Bhattacharjee, Rakesh Das

Hyperspectral image (HSI) classification faces critical challenges, including high spectral dimensionality, complex spectral-spatial correlations, and limited training samples with severe class imbalance. While CNNs excel at local feature extraction and transformers capture long-range dependencies, their isolated application yields suboptimal results due to

Matéo Ghezal

We show that $\mathcal{C}^{\infty}$ local diffeomorphisms of closed surfaces whose topological entropy is larger than the logarithm of their degree admit a finite number of ergodic measures of maximal entropy. To do this, we construct families of rectangles, with a nice geometry, displaying a Markov property. We then analyze the behavior of the iterates of u

Baolong Bi, Shenghua Liu, Yiwei Wang, Siqian Tong

Recent advances in reinforcement learning (RL) have significantly improved the complex reasoning capabilities of large language models (LLMs). Despite these successes, existing methods mainly focus on single-domain RL (e.g., mathematics) with verifiable rewards (RLVR), and their reliance on purely online RL frameworks restricts the exploration space, thereby

Xueyan Hu, Jerome P. Reiter

In many settings, a data curator links records from two files to produce datasets that are shared with secondary analysts. Analysts use the linked files to estimate models of interest, such as regressions. Such two-stage approaches do not necessarily account for uncertainty in model parameters that results from uncertainty in the linkages. Further, they do n

Sajjad Pakdamansavoji, Kumar Vaibhav Jha, Baher Abdulhai, James H Elder

Accurate turning movement counts at intersections are important for signal control, traffic management and urban planning. Computer vision systems for automatic turning movement counts typically rely on visual analysis in the image plane of an infrastructure camera. Here we explore potential advantages of back-projecting vehicles detected in one or more infr

Eric Gilbertson, Richard Hensley, Andrew Kirmse, Kyle Bretherton

Mountainous terrain is increasingly being measured and mapped by airplane-based LiDAR (Light Detection and Ranging) techniques, but the accuracy of these measurements in such topographically variable terrain is not well understood. For this study we measured 179 mountain summits with differential GNSS static surveys and compared summit elevation and location

Katarzyna Fojcik, Renaldas Zioma, Jogundas Armaitis

Efficient machine learning deployment requires models that account for hardware constraints. Because binary logic gates are the fundamental primitives of digital hardware, models built directly from logic operations offer a promising path toward highly energy-efficient computation. Recent work has shown that networks of binary logic gates can be trained with

Felix Eder, Catherine Witteveen, Enrico Giannini, Fabian O. von Rohr

Layered delafossite-type magnetic materials, such as KCrSe$_2$, are promising platforms for studying magnetic systems and potential frustration on triangular lattices. Synthesis, structure-type control, and off-stoichiometries remain major challenges in the investigation of these delafossite-type magnets. Starting from the same self-flux composition (K:Cr:Se

Eric Gilbertson, Matthew Gilbertson

Not all nations on earth have previously been surveyed accurately enough to know for certain which peak is the national highpoint, the highest peak in the country. Knowledge of these peaks is important for understanding the physical geography of these countries in terms of natural resource availability, watershed management, and tourism potential. For this s

Mehrzad Ajoodanian

We study projectively flat holomorphic vector bundles over Riemann surfaces. To each such bundle, we naturally assign a Wronskian line bundle. The main idea is a notion of the division of two meromorphic sections. Abel's identity is interpreted as the first Chern class of the Wronskian line bundle.

Alexander Hammerl, Ravi Seshadri, Thomas Kjær Rasmussen, Otto Anker Nielsen

This paper presents a hierarchical longitudinal control architecture for autonomous truck platoons that jointly addresses safety, string stability, and economic efficiency. The framework integrates a high-rate safety projection filter, a spacing-regulation layer based on a lag-aware proportional-integral-derivative (PID) controller, and a slow-timescale econ

Marios Kalomenopoulos, Riccardo Barbieri, Sadegh Khochfar, Jonathan Gair

Gravitational waves (GWs) offer an alternative way to measure the Hubble parameter. The optimal technique, the ``bright siren'' approach, requires the identification of an electromagnetic counterpart. However, a significant fraction of gravitational waves signals will not have counterparts. Such events can still constrain the Hubble parameter $H_0$ via stati

Wei Jin, Lang Lang, Amanda B. Spence, Leah H. Rubin

Learning causality from observational data has received increasing interest across various scientific fields. However, most existing methods assume the absence of latent confounders and restrict the underlying causal graph to be acyclic, assumptions that are often violated in many real-world applications. In this paper, we address these challenges by proposi

Eric Gilbertsona, Mooketsi Segobye, Yashon Ouma, Boipuso Nkwae

Botswana has not previously been surveyed with sufficient accuracy to determine the highest peak in the country. Otse Hill and Monalanong Hill have been identified as the highest peaks, but there was uncertainty on which is highest. For this study, ground surveys were conducted on each of these peaks using an Abney level to identify the peak location and a G

Sepehr Kazemi Ranjbar, Kumail Alhamoud, Marzyeh Ghassemi

Vision-Language Models (VLMs) struggle with negation. Given a prompt like "retrieve (or generate) a street scene without pedestrians," they often fail to respect the "not." Existing methods address this limitation by fine-tuning on large negation datasets, but such retraining often compromises the model's zero-shot performance on affirmative prompts. We show

Arman Pourghorban, Dipankar Maity

We consider a variant of the target defense problem where a single defender is tasked to capture a sequence of incoming intruders. Both the defender and the intruders have non-holonomic dynamics. The intruders' objective is to breach the target perimeter without being captured by the defender, while the defender's goal is to capture as many intruders as poss

Paul Szeptycki, Hongwei Wen

We consider sets of reals $X$ endowed with the Sorgenfrey lower limit topology denoted $X[\leq]$. Przymusi\'nski proved that if $X$ is a $Q$-set then $(X[\leq])^2$ is normal. While the converse is not in general true we consider examples of sets of the reals for which $(X[\leq])^2$ is normal or just pseudo-normal. For example, if $X$ is a $\lambda$ set, then

Bhaswar B. Bhattacharya, Sanchayan Bhowal, Karambir Das, Laura Eslava

In a multiplex network a common set of nodes is connected through different types of interactions, each represented as a separate graph (layer) within the network. In this paper, we study the asymptotic properties of submultiplexes, the counterparts of subgraphs (motifs) in single-layer networks, in the correlated Erdős-Rényi multiplex model. This is a rando

Chandrasekhar Gokavarapu, Dr D Madhusudhana Rao

Purpose: This study extends the structural theory of finite commutative ternary $\Gamma$-semirings into a computational and categorical framework for explicit classification and constructive reasoning. Methods: Constraint-driven enumeration algorithms are developed to generate all non-isomorphic finite ternary $\Gamma$-semirings satisfying closure, distribut

Mykhailo Vorobiov, Rob Behary, Will Torg, Nicolas DeStefano

We introduce an all-optical quantum-enhanced diagnostic for electric fields in low-temperature plasmas. Trace amounts of rubidium vapor, added to argon plasma, allow us to produce spectrally narrow electric field-sensitive optical resonances via quantum optical effect of Rydberg electromagnetically induced transparency, and to non-invasively measure electric

Xi Ding, Lei Wang, Piotr Koniusz, Yongsheng Gao

Real-world visual data rarely presents as isolated, static instances. Instead, it often evolves gradually over time through variations in pose, lighting, object state, or scene context. However, conventional classifiers are typically trained under the assumption of temporal independence, limiting their ability to capture such dynamics. We propose a simple ye

Muhidin Mohamed, Shubhadeep Mukherjee, Bhavana Baad

Blended learning has become a dominant educational model in higher education in the UK and worldwide, particularly after the COVID-19 pandemic. This is further enriched with accompanying pedagogical changes, such as strengthened asynchronous learning, and the use of AI (from ChatGPT and all other similar tools that followed) and other technologies to aid lea

Luca Corazzini, Elisa Deriu, Marco Guerzoni

Large language models (LLMs) increasingly mediate economic and organisational processes, from automated customer support and recruitment to investment advice and policy analysis. These systems are often assumed to embody rational decision making free from human error; yet they are trained on human language corpora that may embed cognitive and social biases.

Ilias Cherkaoui, Indrakshi Dey

Post-quantum cryptography (PQC) must secure large-scale communication systems against quantum adversaries where classical hardness alone is insufficient and purely quantum schemes remain impractical. Lattice-based key encapsulation mechanisms (KEMs) such as CRYSTALS-Kyber provide efficient quantum-resistant primitives but rely solely on computational hardnes

Yoav Evron, Michal Bar-Asher Siegal, Michael Fire

The recent Artificial Intelligence (AI) revolution has opened transformative possibilities for the humanities, particularly in unlocking the visual-artistic content embedded in historical illuminated manuscripts. While digital archives now offer unprecedented access to these materials, the ability to systematically locate, extract, and analyze illustrations

Md Rashidul Islam, Yiyang Sun

We investigate the modal and non-modal linear amplification mechanisms in the flow over a subsonic open cavity and their subsequent interactions to identify optimal flow perturbations that propagate downstream the cavity and trigger flow transitions. Using both the stationary and time-varying base flows from a Direct Numerical Simulation of a cavity flow at

Zhijun Zeng, Junqing Chen, Zuoqiang Shi

We study an inverse problem for stochastic and quantum dynamical systems in a time-label-free setting, where only unordered density snapshots sampled at unknown times drawn from an observation-time distribution are available. These observations induce a distribution over state densities, from which we seek to recover the parameters of the underlying evolutio

Sebastian Hagedorn, Martín Muñoz, Cristian Riveros, Rodrigo Toro Icarte

Automata learning has many applications in artificial intelligence and software engineering. Central to these applications is the $L^*$ algorithm, introduced by Angluin. The $L^*$ algorithm learns deterministic finite-state automata (DFAs) in polynomial time when provided with a minimally adequate teacher. Unfortunately, the $L^*$ algorithm can only learn DF

Nuno Duarte, Loïc Le Guyader, Viktoria Hinger, Marco Ramilli

Resonant inelastic X-ray scattering (RIXS) is a powerful photon-in, photon-out spectroscopy technique for probing electronic, magnetic, and lattice excitations in matter. Time-resolved RIXS extends this capability through a stroboscopic optical pump-probe scheme to characterize the time evolution of the photoexcitation and subsequent relaxation dynamics of a

Bhawna Mukhija, Amit Kashi

Massive stars can exhibit giant eruptions with high mass loss shortly before their explosion as a core-collapse Supernova. These multiple giant eruptions (MGEs) may have a commutative effect that brings the star to a different state, possible one that favors the explosion. To address this problem, we evolve a 100 solar mass star and initiate a series of thre

Jie Ou, Shuaihong Jiang, Yingjun Du, Cees G. M. Snoek

Parameter-efficient fine-tuning (PEFT) methods, such as LoRA, DoRA, and HiRA, enable lightweight adaptation of large pre-trained models via low-rank updates. However, existing PEFT approaches apply static, input-agnostic updates to all tokens, disregarding the varying importance and difficulty of different inputs. This uniform treatment can lead to overfitti

Hideyasu Yamashita

Unlike QFT in Minkowski spacetime (QFTM), QFT in curved spacetime (QFTCS) suffers from a conceptual obscurity on the empirical (experimentally verifiable/falsifiable) laws. We propose to employ the notion of prior conditional probabilities to describe a part of the empirical laws of QFTCS. This is interpreted as a quantum conditional probability without no i

Morteza Shokrani, Felix Maximilian Graf, Anton Kompatscher, Dennis Derewjanko

The charge carrier localization length {\alpha} is a crucial, yet often ignored parameter of conjugated polymers that exponentially influences electronic conductivity. Here, we argue it is a unique proxy of the energy landscape as determined by sample morphology and experienced by mobile charges. To determine {\alpha}, we use that in disordered organic semic

Anton Kompatscher, Morteza Shokrani, Johanna Feurstein, Martijn Kemerink

One of the characteristics of disordered semiconductors is the slow thermalization of charge carriers after excitation due to photoabsorption or high electric fields. An elegant way to capture the effects of the latter on the conductivity is through a field-dependent effective electronic temperature T_eff that can significantly exceed that of the lattice. De

Darvin Yi, Teng Liu, Mattie Terzolo, Lance Hasson

As large language model (LLM) agents increasingly undertake digital work, reliable frameworks are needed to evaluate their real-world competence, adaptability, and capacity for human collaboration. Existing benchmarks remain largely static, synthetic, or domain-limited, providing limited insight into how agents perform in dynamic, economically meaningful env

Zhizhen Li, Xuanhao Luo, Xueren Ge, Longyu Zhou

Large AI models have been widely adopted in wireless communications for channel modeling, beamforming, and resource optimization. However, most existing efforts remain limited to single-modality inputs and channel-specific objec- tives, overlooking the broader potential of large foundation models for unified wireless sensing. To bridge this gap, we propose M

Qifeng Chen, Jiarun Liu, Rengan Xie, Tao Tang

Recent GS-based rendering has made significant progress for LiDAR, surpassing Neural Radiance Fields (NeRF) in both quality and speed. However, these methods exhibit artifacts in extrapolated novel view synthesis due to the incomplete reconstruction from single traversal scans. To address this limitation, we present LiDAR-GS++, a LiDAR Gaussian Splatting rec

Esmaeil Rostami

We develop the theory of residuated lattices by introducing and studying several new types of filters and related concepts, including semi-simple filters, essential filters, the socle of a filter, and independent families of filters. Our primary goal is to understand the inner structure of residuated lattices by analyzing these new objects. First, we establi

Zakhar Kabluchko, Boris Khoruzhenko, Alexander Marynych

This paper investigates asymptotic distribution of complex zeros of random polynomials $P_n(z):=\sum_{k=0}^{n}b(k)\xi_k z^k$, as $n\to\infty$, where $b$ is a regularly varying function at infinity with index $\alpha\in \mathbb{R}$ and $(\xi_k)_{k\geq 0}$ is a sequence of independent copies of a complex-valued random variable $\xi$. The limiting distribution

Chi Liu, Jincheng Liu, Congcong Zhu, Minghao Wang

Developing Medical AI relies on large datasets and easily suffers from data scarcity. Generative data augmentation (GDA) using AI generative models offers a solution to synthesize realistic medical images. However, the bias in GDA is often underestimated in medical domains, with concerns about the risk of introducing detrimental features generated by AI and

Naoya Sugiura, Kosuke Yamada, Yasuhiro Ogawa, Katsuhiko Toyama

LLMs have achieved performance that surpasses humans in many NLP tasks. However, it remains unclear whether problems that are difficult for humans are also difficult for LLMs. This study investigates how the difficulty of quizzes in a buzzer setting differs between LLMs and humans. Specifically, we first collect Japanese quiz data including questions, answer

Kwangho Choiy, Shiv Prakash Patel

Let $F$ be a finite field, and let $\mathbb{E}$ be either a quadratic field extension $E/F$ or the split algebra $F \oplus F$. We study distinguished representations of $\rm{SL}_{2n}(F)$ by the subgroup $H_{\flat} := \rm{SL}_{2n}(F) \cap \rm{GL}_{n}(\mathbb{E})$, which is a variation of the work of Anandavardhanan and Prasad on distinguished representations

Averkios Averkiou, Monica Musso

We revisit the vortex filament conjecture for three-dimensional inviscid and incompressible Euler flows with helical symmetry and no swirl. Using gluing arguments, we provide the first construction of a smooth helical vortex filament in the whole space $\mathbb{R}^3$ whose cross-sectional vorticity is compactly supported in $\mathbb{R}^2$ for all times. The

Hasini Jayathilaka

Prompt injection attacks are an emerging threat to large language models (LLMs), enabling malicious users to manipulate outputs through carefully designed inputs. Existing detection approaches often require centralizing prompt data, creating significant privacy risks. This paper proposes a privacy-preserving prompt injection detection framework based on fede

Bodhisatwa Chatterjee, Drew Zagieboylo, Sana Damani, Siva Hari

Large Language Models (LLMs) are increasingly used to automatically generate optimized CUDA kernels, substantially improving developer productivity. However, despite rapid generation, these kernels often contain subtle correctness bugs and lack formal safety guarantees. Runtime testing is inherently unreliable - limited input coverage and reward hacking can

Alberto Enciso, Antonio J. Fernández, David Ruiz

For any positive integer $k$, we prove the existence of nontrivial $C^k$-smooth uniformly rotating solutions to the 2D incompressible Euler equations with compact spatial support. These solutions, which can be chosen to be small perturbations of radial flows, are the first example of smooth rotating flows with finite energy which are not locally radial. We a

Bohan Li, Wenyuan Li, Kenneth Tsz Hin Ng, Sheung Chi Phillip Yam

A mutual insurance company (MIC) is a type of consumer cooperative owned by its policyholders. By purchasing insurance from an MIC, policyholders effectively become member-owners of the company and are entitled to a share of the surplus, which is determined by their own collective claims and premium contributions. This sharing mechanism creates an interactiv

Andrea Bertogalli, Giacomo Boracchi, Luca Magri

We present a novel multi-modal extrinsic calibration framework designed to simultaneously estimate the relative poses between event cameras, LiDARs, and RGB cameras, with particular focus on the challenging event camera calibration. Core of our approach is a novel 3D calibration target, specifically designed and constructed to be concurrently perceived by al

Marius Bargo, Yacouba Simpore

This paper presents an age-structured, non-autonomous logistic model describing the aquatic and adult stages of the dynamics of malaria-vector mosquitoes. We propose a biological control strategy targeting the aquatic compartment and implement a tracking control for its emergence. A feedback control law guarantees stabilization of the emergent population den

Yihan Dai, Sijie Liang, Haotian Xu, Peichu Xie

Large language models (LLMs) can generate executable code from natural language descriptions, but the resulting programs frequently contain bugs due to hallucinations. In the absence of formal specifications, existing approaches attempt to assess correctness using LLM-generated proxies such as tests or auto-formalized specifications. However, these proxies a

Duilio De Santis, Marios H. Michael, Sambuddha Chattopadhyay, Andrea Cavalleri

Strong optical drives have been shown to induce transient superconducting-like response in materials above their equilibrium $T_c$. Many of these materials already exhibit short-range superconducting correlations in equilibrium. This motivates the question: can external driving enhance coherence in systems with superconducting correlations but no long-range

Khyati Kiyawat, Zhenxing Fan, Yasas Seneviratne, Morteza Baradaran

Large Language Models (LLMs) are becoming increasingly data-intensive due to growing model sizes, and they are becoming memory-bound as the context length and, consequently, the key-value (KV) cache size increase. Inference, particularly the decoding phase, is dominated by memory-bound GEMV or flat GEMM operations with low operational intensity (OI), making

Minu Kim, Ji Sub Um, Hoirin Kim

Lexical tone is central to many languages but remains underexplored in self-supervised learning (SSL) speech models, especially beyond Mandarin. We study four languages with complex and diverse tone systems (Burmese, Thai, Lao, and Vietnamese) to ask how far such models "listen" for tone and how transfer operates in low-resource conditions. As a baseline ref

Ebrahim Ghorbani, Jana Katharina Nickel, Florian Reich

Menger's theorem - the maximum number of vertex-disjoint $X$-$Y$ paths is equal to the minimum size of an $X$-$Y$ separator - is generally not true in bidirected graphs. We prove that Menger's theorem holds true if we take the nontrivial $X$-$X$ paths and the nontrivial $Y$-$Y$ paths into account.

Oleg Ogievetsky, Pavel Pyatov

For a family of the orthogonal $O(k)$ type Quantum Matrix algebras we establish an analogue of the Cayley--Hamilton theorem. The form of the Cayley-Hamilton identity is different in three cases. First, the cases of odd ($k=2\ell -1$) and even ($k=2\ell$) heights are different. Second, for even height orthogonal Quantum Matrix algebra we derive two versions o

Christian Kuehn, Fergal Murphy

We study when co-evolving (or adaptive) higher-order networks defined on directed hypergraphs admit a simplicial description. Binary and triadic couplings are modelled by time-dependent weight tensors. Using representation theory of the symmetric group $S_k$, we decompose these tensors into fully symmetric, fully antisymmetric, and mixed isotypic components,

Ivan Zakazov, Berke Argin, Oussama Gabouj, Kamel Charaf

Motivated by the high costs of using black-box Large Language Models (LLMs), we introduce a novel prompt compression paradigm, under which we use smaller LLMs to compress inputs for the larger ones. We present the first comprehensive LLM-as-a-compressor benchmark spanning 25 open- and closed-source models, which reveals significant disparity in models' compr

Shuochen Chang, Xiaofeng Zhang, Qingyang Liu, Li Niu

Diffusion-based multimodal large language models (Diffusion MLLMs) have recently demonstrated impressive non-autoregressive generative capabilities across vision-and-language tasks. However, Diffusion MLLMs exhibit substantially slower inference than autoregressive models: Each denoising step employs full bidirectional self-attention over the entire sequence

Shubhransh Singhvi, Saransh Chopra, K. V. Rashmi

Recent advances in erasure coding for distributed storage systems have demonstrated that adapting redundancy to varying disk failure rates can lead to substantial storage savings. Such adaptation requires code conversion, wherein data encoded under an initial $[k^I + r^I, k^I]$ code is transformed into data encoded under a final $[k^F + r^F, k^F]$ code - an

Mingqi Wu, Qiang Sun, Yi Yang

High-dimensional data often contain low-dimensional signals obscured by structured background noise, which limits the effectiveness of standard PCA. Motivated by contrastive learning, we address the problem of recovering shared signal subspaces from positive pairs, paired observations sharing the same signal but differing in background. Our baseline, PCA+, u