October 2025 arXiv papers — page 94

Panagiotis Charalampopoulos, Tomasz Kociumaka, Philip Wellnitz

In Pattern Matching with Weighted Edits (PMWED), we are given a pattern $P$ of length $m$, a text $T$ of length $n$, a positive threshold $k$, and oracle access to a weight function that specifies the costs of edits (depending on the involved characters, and normalized so that the cost of each edit is at least $1$). The goal is to compute the starting positi

M. R. Austin, L. Carmichael, D. McArthur, E. Milton

The current Timing System at Fermilab has been around for 40 years and currently relies on 7 CAMAC crates and over 100 CAMAC cards to produce the Tevatron Clock (TCLK). Thanks to the ingenuity of those before us, this has allowed Fermilab the flexibility to change the timing and Events for its accelerator as beamlines and projects have changed over the years

Nikita Ustimenko, Andrey B. Evlyukhin, Vicky Kyrimi, Alexander V. Kildishev

A referential example of a physical system that supports bound states in the continuum (BICs) with an infinite quality factor ($Q$ factor) is a metasurface of discrete scatterers (resonators), whose response can be significantly modified by exploiting lattice interactions. In this work, we explore the multipole-interference mechanism for realizing accidental

Katharina Kormanna, Giorgia Testolina

Central configurations play a fundamental role in the Newtonian $n$-body problem, as they give rise to motions in which the configuration evolves while preserving its shape up to rotation and scaling. These include relative equilibria, where the configuration rigidly rotates about the center of mass and each body moves along a circular orbit. For $d\le3$, su

William Zhang, Wan Shen Lim, Andrew Pavlo

Tuning database management systems (DBMSs) is challenging due to trillions of possible configurations and evolving workloads. Recent advances in tuning have led to breakthroughs in optimizing over the possible configurations. However, due to their design and inability to leverage query-level historical insights, existing automated tuners struggle to adapt an

K. R. Prathyusha, Paulami Sarkar, Justin Xu, Saad Bhamla

Living organisms employ diverse strategies to navigate confined environments. Inspired by translocation observations on California blackworms (\textit{Lumbriculus variegatus}), we combine biological experiments and active-polymer simulations to examine how confinement and stiffness govern translocation. Active filaments translocate fastest when the channel w

Casey Y. Lam, Joshua D. Simon, Kareem El-Badry, Howard Isaacson

Stellar-mass black holes (BHs) above $30 M_\odot$ are predicted to form from low-metallicity progenitors, but direct detections of such systems in the Milky Way remain scarce. Motivated by the recent discovery of Gaia BH3, a $33 M_\odot$ BH with a very metal-poor giant companion, we conduct a systematic search for additional systems. Approximately 900 candid

Martí Berenguer Mimó

This thesis studies the non-equilibrium dynamics of strongly coupled quantum systems within the framework of the AdS/CFT correspondence, with particular emphasis on periodically driven (Floquet) systems. The first part focuses on top-down holographic constructions based on D3/D5 and D3/D7 brane intersections subjected to time-dependent external fields. In th

Lars Niedermeier, Vyom Shah, Jeffrey L. Krichmar

Spiking Neural Networks (SNNs) have sparse, event driven processing that can leverage neuromorphic applications. In this work, we introduce a multi-threading kernel that enables neuromorphic applications running at the edge, meaning they process sensory input directly and without any up-link to or dependency on a cloud service. The kernel shows speed-up gain

Grant McKenzie, Krzysztof Janowicz, Carsten Kessler

Large-scale pre-trained machine learning models have reshaped our understanding of artificial intelligence across numerous domains, including our own field of geography. As with any new technology, trust has taken on an important role in this discussion. In this chapter, we examine the multifaceted concept of trust in foundation models, particularly within a

Noah Kravitz, James Leng

The "pyjama stripe" with parameter $\varepsilon>0$ is the set $E(\varepsilon)$ of all complex numbers $z$ such that the distance from $\Re(z)$ to the nearest integer is at most $\varepsilon$. The Pyjama Problem of Iosevich, Kolountzakis, and Matolcsi asks whether, for every choice of $\varepsilon>0$, it is possible to cover the entire complex plane with fini

Alexandr Grebennikov, Matthew Kwan

How many points can be placed in an $n\times n$ grid so that every (affine) line contains at most $k$ points? We prove that for $n \ge k \ge 10^{37}$ the maximum number of points is exactly $kn$. Our proof builds on the recent work of Kov\'acs, Nagy, and Szab\'o (who proved an analogous result when $k$ is at least about $\sqrt{n \log n}$), incorporating idea

Stewart Slocum, Julian Minder, Clément Dumas, Henry Sleight

Knowledge editing techniques promise to implant new factual knowledge into large language models (LLMs). But do LLMs really believe these facts? We develop a framework to measure belief depth and use it to evaluate the success of knowledge editing techniques. We operationalize belief depth as the extent to which implanted knowledge 1) generalizes to related

Zhiming Lin

Multi turn intent understanding is central to task oriented chatbots, yet real deployments face tight token budgets and noisy contexts, and most retrieval pipelines emphasize relevance while overlooking set level diversity and confounds such as more context or exemplar order. We ask whether retrieval diversity, rather than longer prompts, systematically impr

Alejandro Mir, Jorge Alda, Siannah Penaranda

Discrepancies between experimental measurements and Standard Model predictions in $B$-meson decays, especially in lepton flavor universality ratios like $R_{D^{(*)}}$, $R_{J/\psi}$ and branching ratios for processes like $B\to K^+\nu\bar\nu$, suggest possible new physics (NP). In this study, we use an effective field theory framework, assuming NP effects onl

Navid Reyhanian, Reza Ghaderi Zefreh, Parisa Ramezani, Emil Bjornson

This paper studies a reconfigurable intelligent surface (RIS)-assisted cell-free massive multiple-input multiple-output (CF-mMIMO) system with multiple RISs. Joint design of transmit precoding, RIS coefficients, and receive combining is investigated for uplink sum-rate maximization under in-phase and quadrature phase imbalance (IQI) at user equipments (UEs)

Shunhua Jiang, Michael Kapralov, Lawrence Li, Aaron Sidford

In this paper we consider generalized flow problems where there is an $m$-edge $n$-node directed graph $G = (V,E)$ and each edge $e \in E$ has a loss factor $\gamma_e >0$ governing whether the flow is increased or decreased as it crosses edge $e$. We provide a randomized $\tilde{O}( (m + n^{1.5}) \cdot \mathrm{polylog}(\frac{W}{\delta}))$ time algorithm for

Timur Ismagilov, Shakaiba Majeed, Michael Milford, Tan Viet Tuyen Nguyen

We address multi-reference visual place recognition (VPR), where reference sets captured under varying conditions are used to improve localisation performance. While deep learning with large-scale training improves robustness, increasing data diversity and model complexity incur extensive computational cost during training and deployment. Descriptor-level fu

Yifeng Liu, Angela Yuan, Quanquan Gu

Matrix-based preconditioned optimizers, such as Muon, have recently been shown to be more efficient than scalar-based optimizers for training large-scale neural networks, including large language models (LLMs). Recent benchmark studies of LLM pretraining optimizers have demonstrated that variance-reduction techniques such as MARS can substantially speed up t

Fabian Jaensch, Giuseppe Caire, Begüm Demir

In 5G, beam training consists of the efficient association of users to beams for a given beamforming codebook used at the base station and the given propagation environment in the cell. We propose a convolutional neural network approach that leverages the position of the base station and geospatial data to predict beam distributions for all user locations si

Zachary Abel, Erik D. Demaine, Martin L. Demaine, Sarah Eisenstat

We exactly settle the complexity of graph realization, graph rigidity, and graph global rigidity as applied to three types of graphs: "globally noncrossing" graphs, which avoid crossings in all of their configurations; matchstick graphs, with unit-length edges and where only noncrossing configurations are considered; and unrestricted graphs (crossings allowe

Luochen Zhao

Let $E$ be an elliptic curve having CM by the ring of integers of an imaginary quadratic field $K$ in which $p$ splits. Following Lichtenbaum, the Bernoulli--Hurwitz numbers of $E$ (i.e., values of Eisenstein series evaluated at $E$ up to normalization) admit integral representations given by a $p$-adic measure constructed from an elliptic function. We show

Jun Yan

The celebrated result of Koml\'os, S\'ark\"ozy, and Szemer\'edi states that for any $\varepsilon>0$, there exists $0<c<1$, such that for all sufficiently large $n$, every $n$-vertex graph $G$ with $\delta(G)\geq(1/2+\varepsilon)n$ contains every $n$-vertex tree with maximum degree at most $cn/\log n$. This is best possible up to the value of $c$. In this pap

Omer Bahadir Eryilmaz, Cihan Katar, Max A. Little

Recurrent signals give rise to trajectories that repeatedly return close to earlier states in state space. Many analysis methods therefore require a principled notion of similarity between states. In practice, a recurrence threshold sets the scale of the neighbourhood used to define when two states are considered close. Close returns can also support topolog

Navjot Singh, Edgar Solomonik, Xiaoye Sherry Li, Yang Liu

This paper presents low-complexity tensor completion algorithms and their efficient implementation to reconstruct highly oscillatory operators discretized as $n\times n$ matrices. The underlying tensor decomposition is based on the reshaping of the input matrix and its butterfly decomposition into an order $O (\log n)$ tensor. The reshaping of the input matr

Tong Chen, Akari Asai, Luke Zettlemoyer, Hannaneh Hajishirzi

Language models often generate factually incorrect information unsupported by their training data, a phenomenon known as extrinsic hallucination. Existing mitigation approaches often degrade performance on open-ended generation and downstream tasks, limiting their practical utility. We propose an online reinforcement learning method using a novel binary retr

Aaron Appelle, Jerome P. Lynch

Recent high-performing image-to-video (I2V) models based on variants of the diffusion transformer (DiT) have displayed remarkable inherent world-modeling capabilities by virtue of training on large scale video datasets. We investigate whether these models can generate realistic pedestrian movement patterns in crowded public scenes. Our framework conditions I

Valentin Durupt, Fabio Maltoni, Olivier Mattelaer

We present a fully automated framework to compute production spin-density matrices for generic collider processes at tree level within \textsc{MadGraph5\_aMC@NLO}. The method assembles helicity amplitudes into event-by-event production matrices. These are written to the LHE file in a compact form, together with run metadata, enabling direct post-processing o

Alessandro Giacchetto, Pronobesh Maity, Edward A. Mazenc

We show certain correlators in generic one-matrix models define a notion of ``discrete'' volumes of the moduli space of Riemann surfaces, generalizing the connection between random matrices and JT gravity. We prove they obey a discrete, Mirzakhani-like recursion relation. Their fundamental discreteness crucially relies upon studying these matrix integrals aw

Ege Beyazit, KL Navaneet, Prashant Mathur, Roi Blanco

Black-box Large Language Models (LLMs) provide practical and accessible alternatives to other machine learning methods, as they require minimal labeled data and machine learning expertise to develop solutions for various decision making problems. However, for applications that need operating with constraints on specific metrics (e.g., precision $\geq$ 95%),

Md. Faiyaz Abdullah Sayeedi, Md. Sadman Haque, Zobaer Ibn Razzaque, Robiul Awoul Robin

With the increasing integration of Artificial Intelligence (AI) in academic problem solving, university students frequently alternate between traditional search engines like Google and large language models (LLMs) for information retrieval. This study explores students' perceptions of both tools, emphasizing usability, efficiency, and their integration into

Haozhen Zhang, Tao Feng, Pengrui Han, Jiaxuan You

Large Language Models (LLMs) have recently achieved remarkable performance in long-context understanding. However, current long-context LLM benchmarks are limited by rigid context length, labor-intensive annotation, and the pressing challenge of label leakage issues during LLM training. Therefore, we propose \textsc{AcademicEval}, a live benchmark for evalua

Matheus Ramos Parracho

Automated signature verification is a critical biometric technique used in banking, identity authentication, and legal documentation. Despite the notable progress achieved by deep learning methods, most approaches in offline signature verification still struggle to generalize across datasets, as variations in handwriting styles and acquisition protocols ofte

J. Tian, S. Singh, B. W. Stappers, J. D. Turner

We present the discovery of 30 new Galactic sources from the MeerTRAP project, a commensal fast radio transient search programme using the MeerKAT telescope. These sources were all identified via a single pulse search. Most of them are likely to be rotating radio transients (RRATs) given their low pulse rates. Using data captured in our transient buffer we h

Yaning Pan, Qianqian Xie, Guohui Zhang, Zekun Wang

The recent development of Multimodal Large Language Models (MLLMs) has significantly advanced AI's ability to understand visual modalities. However, existing evaluation benchmarks remain limited to single-turn question answering, overlooking the complexity of multi-turn dialogues in real-world scenarios. To bridge this gap, we introduce MT-Video-Bench, a hol

Jesús D. Petro-Ramos, David J. Ruiz-Morales, D. Sierra-Porta

We investigate graph-based representations of astronomical light curves for transient classification on a quality-controlled, class-balanced subset of the MANTRA benchmark (minimum coverage N_min=100 epochs; N=1705 objects after filtering and Non-Tr. subsampling). Each series is mapped to three visibility-graph views -- horizontal (HVG), directed (DHVG), and

Nanda Kumar Rengarajan, Jun Yan, Chun Wang

Named Entity Recognition (NER) is a critical task that requires substantial annotated data, making it challenging in low-resource scenarios where label acquisition is expensive. While zero-shot and instruction-tuned approaches have made progress, they often fail to generalize to domain-specific entities and do not effectively utilize limited available data.

Zhiqiang Teng, Tingting Chen, Beibei Lin, Zifeng Yuan

3D Gaussian Splatting (3DGS) under raindrop conditions suffers from severe occlusions and optical distortions caused by raindrop contamination on the camera lens, substantially degrading reconstruction quality. Existing benchmarks typically evaluate 3DGS using synthetic raindrop images with known camera poses (constrained images), assuming ideal conditions.

Senhao Duan

In this paper, we prove the existence of a singular standing sphere blow-up solution for the nonlinear heat equation with radial symmetry. This solution develops a finite-time singularity on a fixed-radius sphere and exhibits a flat blow-up profile. Our construction refines the method developed by Merle and Zaag \cite{MZJEMS24} which reduces the infinite-dim

Young Kun Ko

We prove a general translation theorem for converting one-way communication lower bounds over a product distribution to dynamic cell-probe lower bounds. Specifically, we consider a class of problems considered in [Pat10] where: 1. $S_1, \ldots, S_m \in \{0, 1\}^n$ are given and publicly known. 2. $T \in \{0, 1\}^n$ is a sequence of updates, each taking $t_u$

Suqiang Ma, Subhadeep Sengupta, Yao Lee, Beikang Gu

Circulating blood cell clusters (CCCs) containing red blood cells (RBCs), white blood cells(WBCs), and platelets are significant biomarkers linked to conditions like thrombosis, infection, and inflammation. Flow cytometry, paired with fluorescence staining, is commonly used to analyze these cell clusters, revealing cell morphology and protein profiles. While

Hanxu Hu, Xingxing Zhang, Jannis Vamvas, Rico Sennrich

Large Language Models have achieved strong performance on reasoning tasks, solving competition-level coding and math problems. However, their scalability is limited by human-labeled datasets and the lack of large-scale, challenging coding problem training data. Existing competitive coding datasets contain only thousands to tens of thousands of problems. Prev

Vinícius Pereira da Silva Oliveira, Danilo da Silva Borges, Erick de Moraes Franklin

Fluidized beds consist of solid particles suspended in a tube by an ascending fluid. In liquids, it is not rare that particles adhere to each other, decreasing the solid-liquid contact area and the ratio between the tube and grain diameters, deteriorating fluidization. We inquire into this problem by carrying out experiments with trios of spheres fluidized b

Phuong Q. Dao, Mark Roantree, Vuong M. Ngo

Multimodal Sentiment Analysis (MSA) seeks to understand human emotions by jointly analyzing data from multiple modalities typically text and images offering a richer and more accurate interpretation than unimodal approaches. In this paper, we first propose BERT-ViT-EF, a novel model that combines powerful Transformer-based encoders BERT for textual input and

Mangsura Kabir Oni, Tabia Tanzin Prama

Machine Translation (MT) has advanced from rule-based and statistical methods to neural approaches based on the Transformer architecture. While these methods have achieved impressive results for high-resource languages, low-resource varieties such as Sylheti remain underexplored. In this work, we investigate Bengali-to-Sylheti translation by fine-tuning mult

Atticus McWhorter, Daryl DeFord

Novel Markov Chain Monte Carlo (MCMC) methods have enabled the generation of large ensembles of redistricting plans through graph partitioning. However, existing algorithms such as Reversible Recombination (RevReCom) and Metropolized Forest Recombination (MFR) are constrained to sampling from distributions related to spanning trees. We introduce the marked e

E. Bortchagovsky, A. V. Chumak, V. Lozovski

Brillouin light scattering (BLS) is a key technique in studying magnonic systems, but its sensitivity is often limited. While nanoplasmonic systems can enhance BLS through near-field effects, we propose a novel approach for additional amplification. In this conceptual paper, we show how to actively supply energy to a surface collective electromagnetic resona

Frederik J. Zuiderveen Borgesius, Judith Möller, Sanne Kruikemeier, Ronan Ó Fathaigh

Online political microtargeting involves monitoring people's online behaviour, and using the collected data, sometimes enriched with other data, to show people-targeted political advertisements. Online political microtargeting is widely used in the US; Europe may not be far behind. This paper maps microtargeting's promises and threats to democracy. For examp

Frederik Zuiderveen Borgesius

Artificial intelligence (AI) has a huge impact on our personal lives and also on our democratic society as a whole. While AI offers vast opportunities for the benefit of people, its potential to embed and perpetuate bias and discrimination remains one of the most pressing challenges deriving from its increasing use. This new study, which was prepared by Prof

Frederik Zuiderveen Borgesius

Information about millions of people is collected for behavioural targeting, a type of marketing that involves tracking people's online behaviour for targeted advertising. It is hotly debated whether data protection law applies to behavioural targeting. Many behavioural targeting companies say that, as long as they do not tie names to data they hold about in

R. Nasery, P. Varghese, S. Raman, M. Guran

The PIP-II linac is an international collaboration project with in kind contributions of key subsystems from multiple countries including India (DAE). In the research and development phase of the project, the LLRF and resonance control systems were jointly developed by BARC and Fermilab and were delivered to Fermilab for testing and validation. Initial testi

Ana Enériz Janeiro, Karina Pitombeira Pereira, Julio Mayol, Javier Crespo

AI is transforming medical practice and redefining the competencies that future healthcare professionals need to master. Despite international recommendations, the integration of AI into Medicine curricula in Spain had not been systematically evaluated until now. A cross-sectional study (July-September 2025) including Spanish universities offering the offici

Omar Alvarado-Garduño, Jesús González

For positive integers $n$, $p$ and $q$ with $pq-n>0$, let $UC(n,p\times q)$ denote the configuration space of $n$ unlabelled hard unit squares in the rectangle $[0,p]\times[0,q]$, and let $B_n(p\times q)$ denote the corresponding fundamental group. It is known that, as $pq-n$ becomes large, $UC(n,p\times q)$ starts capturing homotopical properties of the cla

Yu Tao, Li Lei, Jingyi Liu, Binbin Wu

Under cold compression, hydrogen bonding was considered to dominate intermolecular interaction during the ionization of ammonia. Here, we provide experimental and theoretical evidence of intermolecular HH coupling in dense ammonia. Ab initio molecular dynamics simulations (AIMD) reveal an increasing degree of proton disorder in ammonia with increasing pressu

Dayan Pan, Zhaoyang Fu, Jingyuan Wang, Xiao Han

Large Language Models (LLMs) possess remarkable generalization capabilities but struggle with multi-task adaptation, particularly in balancing knowledge retention with task-specific specialization. Conventional fine-tuning methods suffer from catastrophic forgetting and substantial resource consumption, while existing parameter-efficient methods perform subo

Mhd Adnan Albani, Riad Sonbol

Parkinson's disease (PD) is a neurodegenerative disease affecting about 1% of people over the age of 60, causing motor impairments that impede hand coordination activities such as writing and drawing. Many approaches have tried to support early detection of Parkinson's disease based on hand-drawn images; however, we identified two major limitations in the re

Roy Magen

In this article, we introduce fundamental notions and results about pullback formalisms, building on work of Drew-Gallauer. Our main application is producing a pullback formalism $\mathbf{SH}^{\mathrm{hol}}$ that encodes a version of motivic homotopy theory for complex analytic stacks, and establishing some of its properties. The notions introduced in this a

Kristiana Mihali, Dennis Wörthmüller, Pierre Sens

Cell shape changes are largely controlled by the actin cytoskeleton, a dynamic filament network beneath the plasma membrane. Several cell types can form extended free-standing protrusions not supported by an extracellular substrate or matrix, and regulated by proteins that modulate cytoskeletal dynamics in a way sensitive to the curvature of the cell membran

Aleksandr Oganov, Ilya Bykov, Eva Neudachina, Mishan Aliev

While diffusion models achieve state-of-the-art generation quality, they still suffer from computationally expensive sampling. Recent works address this issue with gradient-based optimization methods that distill a few-step ODE diffusion solver from the full sampling process, reducing the number of function evaluations from dozens to just a few. However, the

Hou-Wan Long, Yicheng Song, Zidong Wang, Tianshu Sun

Sponsored search advertising (SSA) requires advertisers to constantly adjust keyword strategies. While bid adjustment and keyword generation are well-studied, keyword pruning-refining keyword sets to enhance campaign performance-remains under-explored. This paper addresses critical inefficiencies in current practices as evidenced by a dataset containing 0.5

Liqun He, Manolis Mavrikis, Mutlu Cukurova

Dialogue plays a crucial role in educational settings, yet existing evaluation methods for educational applications of large language models (LLMs) primarily focus on technical performance or learning outcomes, often neglecting attention to learner-LLM interactions. To narrow this gap, this AIED Doctoral Consortium paper presents an ongoing study employing a

Anjie Liu, Jianhong Wang, Samuel Kaski, Jun Wang

Steering cooperative multi-agent reinforcement learning (MARL) towards desired outcomes is challenging, particularly when the global guidance from a human on the whole multi-agent system is impractical in a large-scale MARL. On the other hand, designing external mechanisms (e.g., intrinsic rewards and human feedback) to coordinate agents mostly relies on emp

Haojie Hu, Yuta Asahina, Shogo Yoshioka, Hiroyuki R. Takahashi

Recent advances in X-ray spectroscopic observation have enabled researchers to reveal distinct clumpy structures in the super-Eddington outflows from the supermassive black hole in PDS 456 (XRISM Collaboration 2025), initiating detailed investigation of fine-scale structures in accretion-driven outflows. In this study, we conduct high-resolution, two-dimensi

Jacopo Tagliabue

We explore AI-driven distributed-systems policy design by combining stochastic code generation from large language models (LLMs) with deterministic verification in a domain-specific simulator. Using a Function-as-a-Service runtime (Bauplan) and its open-source simulator (Eudoxia) as a case study, we frame scheduler design as an iterative generate-and-verify

Juan Gutiérrez, Victor Gutiérrez, Ángel Mora, Silvia Rodriguez

Manual annotation remains the gold standard for high-quality, dense temporal video datasets, yet it is inherently time-consuming. Vision-language models can aid human annotators and expedite this process. We report on the impact of automatic Pre-Annotations from a tuned encoder on a Human-in-the-Loop labeling workflow for video footage. Quantitative analysis

Maximilian H. V. Tillmann, Ban-Sok Shin, Dmitriy Shutin, Armin Dekorsy

Multi-agent systems (MASs) are a promising solution for autonomous exploration tasks in hazardous or remote environments. In such settings, communication among agents is essential to ensure collaborative task execution, yet conventional approaches treat exploration and communication as decoupled subsystems. This work presents an approach that tightly integra

Gökhan Altıner, Jons Bolding, Yiğit Mert Kaplan, Floor Souren

Tandem solar cells offer a promising alternative to exceed the efficiency limits of single-junction silicon photovoltaics, yet they require high-performance recombination junctions that are transparent, passivating, and electrically efficient. Indium tin oxide (ITO), which is conventionally used as a recombination junction material, faces challenges related

Shenglan Sun, Yang Huang, Fangzhou Jiang, Huawei Zhang

The earliest assembly of the Milky Way remains poorly understood, yet the spatial, chemical, and kinematic properties of its most metal-poor stars provide a unique fossil record of its proto-Galaxy phase. Understanding how this ancient component formed is essential for linking near-field Galactic archaeology to high-redshift galaxy evolution. We construct th

Bama Srinivasan

This paper presents a formal framework for sequencing instructions in AI agents, inspired by the Indian philosophical system of Mimamsa. The framework formalizes sequencing mechanisms through action object pairs in three distinct ways: direct assertion (Srutikrama) for temporal precedence, purpose driven sequencing (Arthakrama) for functional dependencies, a

Xihong Su

This dissertation makes three main contributions. First, We identify a new connection between policy gradient and dynamic programming in MMDPs and propose the Coordinate Ascent Dynamic Programming (CADP) algorithm to compute a Markov policy that maximizes the discounted return averaged over the uncertain models. CADP adjusts model weights iteratively to guar

Pau Escofet, Eduard Alarcón, Sergi Abadal, Andrii Semenov

As quantum computers scale toward millions of physical qubits, it becomes essential to robustly encode individual logical qubits to ensure fault tolerance under realistic noise. A high-quality foundational encoding allows future compilation techniques and heuristics to build on optimal or near-optimal layouts, improving scalability and error resilience. In t

Shawn M. Gibford, Mohammad Reza Boskabadi, Christopher J. Savoie, Seyed Soheil Mansouri

Data scarcity and sparsity in bio-manufacturing poses challenges for accurate model development, process monitoring, and optimization. We aim to replicate and capture the complex dynamics of industrial bioprocesses by proposing the use of a Quantum Wasserstein Generative Adversarial Network with Gradient Penalty (QWGAN-GP) to generate synthetic time series d

Xu Zhang, Hao Li, Zhichao Lu

Multimodal Large Language Models (MLLMs) achieve strong reasoning and perception capabilities but are increasingly vulnerable to jailbreak attacks. While existing work focuses on explicit attacks, where malicious content resides in a single modality, recent studies reveal implicit attacks, in which benign text and image inputs jointly express unsafe intent.

Fu-Yun Wang, Han Zhang, Michael Gharbi, Hongsheng Li

We present UniRL-Zero, a unified reinforcement learning (RL) framework that boosts, multimodal language model understanding and reasoning, diffusion model multimedia generation, and their beneficial interaction capabilities within a unified model. Our work defines six scenarios for unified model reinforcement learning, providing systematic baselines for rein

Taichi Liu, Zhenyu Wang, Ruofeng Liu, Guang Wang

Recent advancements in 3D object detection and novel category detection have made significant progress, yet research on learning generalized 3D objectness remains insufficient. In this paper, we delve into learning open-world 3D objectness, which focuses on detecting all objects in a 3D scene, including novel objects unseen during training. Traditional close

Min Cao, Xinyu Zhou, Ding Jiang, Bo Du

Text-to-image person retrieval (TIPR) aims to identify the target person using textual descriptions, facing challenge in modality heterogeneity. Prior works have attempted to address it by developing cross-modal global or local alignment strategies. However, global methods typically overlook fine-grained cross-modal differences, whereas local methods require

Xinwei Zhang, Hu Chen, Zhe Yuan, Sukun Tian

Foundation models for medical image segmentation have achieved remarkable performance. Adaptive fine-tuning of natural image segmentation foundation models is crucial for medical image segmentation tasks. However, some limitations exist in existing fine-tuning methods: 1) insufficient representation of high-level features and 2) the fine-tuning process disru

Sebastiano Battisti, Matteo Pioldi, Alessandro Paghi, Giorgio De Simoni

We present a groundbreaking demonstration of thermal modulation in a field-effect-controllable semiconductor-superconductor hybrid structure, wherein the heating mechanism is exclusively radiative. The architecture comprises two reservoirs separated by $\sim 1$ mm and interconnected via a completely non-galvanic electrical circuit, enabling the transfer of b

T. Muscheid, R. Gartmann, L. E. Ardila-Perez, A. Acevedo-Rentería

The BULLKID-DM experiment aims to detect WIMP-like potential Dark Matter particles with masses below 1 GeV/c^2. Sensing these particles is challenging, as it requires nuclear recoil detectors characterized by high exposure and an energy threshold in the order of 100 eV, thus exceeding the capabilities of conventional semiconductor detectors. BULLKID-DM inten

Yuandong Pu, Le Zhuo, Songhao Han, Jinbo Xing

Image editing has achieved remarkable progress recently. Modern editing models could already follow complex instructions to manipulate the original content. However, beyond completing the editing instructions, the accompanying physical effects are the key to the generation realism. For example, removing an object should also remove its shadow, reflections, a

Luis Ferroni, Roberto Riccardi

Let $P$ be a finite partially ordered set. In a recent series of works, Proudfoot introduced the notion of $Z$-polynomials associated with $P$-kernels, providing a unified framework for various intersection cohomology Poincaré polynomials arising in diverse areas of mathematics. One of the problems posed by Proudfoot was to interpret the $Z$-polynomial in a

Valery Alexeev, Wenfei Liu, Matthias Schütt

Let $M_1$ be the moduli space of the KSBA stable surfaces $X$ of geometric genus $p_g(X)=1$ realizing the minimal possible volume $K_X^2=\frac1{143}$. We show that its reduced part $M_{1,\rm red}$ is a $10$-dimensional projective variety isomorphic to the Baily--Borel compactification $\overline{F}_\Lambda^{\rm BB}$ of the moduli space of $\Lambda$-polarized

Swapnoneel Roy, Asai Asaithambi, Debajyoti Mukhopadhyay

We show that \emph{Sorting by Strip Swaps} (SbSS) is NP-hard by a polynomial reduction of \emph{Block Sorting}. The key idea is a local gadget, a \emph{cage}, that replaces every decreasing adjacency $(a_i,a_{i+1})$ by a guarded triple $a_i,m_i,a_{i+1}$ enclosed by guards $L_i,U_i$, so the only decreasing adjacencies are the two inside the cage. Small \emph{

Jiale Zhao, Cong Liu, Yuxuan Zhang, Chengyue Gong

Determining crystal structures from X-ray diffraction data is fundamental across diverse scientific fields, yet remains a significant challenge when data is limited to low resolution. While recent deep learning models have made breakthroughs in solving the crystallographic phase problem, the resulting low-resolution electron density maps are often ambiguous

Luc Dessart, W. V. Jacobson-Galan

With the ever faster cadence of untargeted surveys of the sky, the supernova (SN) community will capture in the coming years a growing number of shock breakouts in red-supergiant (RSG) stars. Expecting a high frequency of breakouts within circumstellar material (CSM), we have produced an extended, regular and cubic grid of models covering from low- to high-e

Kara D. Lamb, Jerry Y. Harrington, Alfred M. Moyle, Gwenore F. Pokrifka

Depositional ice growth is an important process for cirrus cloud evolution, but the physics of ice growth in atmospheric conditions is still poorly understood. One major challenge in constraining depositional ice growth models against observations is that the early growth rates of ice crystals cannot be directly observed, and proposed models require assumpti

Tengyu Li, Lei Zhang

In this paper, we investigate the Cauchy problem associated with the stochastic hyperbolic Keller-Segel (SHKS) equation featuring multiplicative noises on the torus $\mathbb{T}^d$. First, we establish the local existence and uniqueness of pathwise solutions to the SHKS equation within Sobolev spaces $H^s(\mathbb{T}^d)$ for $s>\frac{d}{2}+1$, under appropriat

Katelyn Lazareno, Christopher Chae, Becky Haight, Shams Jabin

To investigate the scope of ferroelectric behavior in La-substituted BiFeO3 films, LaxBi1-xFeO3 epitaxial films were synthesized using off-axis co-sputtering on SrTiO3(001) and DyScO3(110) substrates with a SrRuO3 bottom electrode layer. A digital-doping deposition method was used to enable precise control and continuous tuning of La concentration in high-qu

Zhiwen Yu, Zhaocheng Liu, Xiaoqing Liu, Huanqiang Zeng

Multimodal learning faces modality imbalance, where dominant modalities suppress weaker ones due to inconsistent convergence rates. Existing static or heuristic methods overlook sample-level variations in prediction bias and fail to isolate low-quality outlier samples. To address this, we propose a novel framework to quantitatively diagnose and dynamically m

Katarzyna Kobalczyk, Zhiyuan Jerry Lin, Benjamin Letham, Zhuokai Zhao

Many real-world optimization problems are guided by complex, subjective preferences that are difficult to express as explicit closed-form objectives. In response, we introduce Language-in-the-Loop Optimization (LILO), a Bayesian optimization (BO) framework that employs a large language model (LLM) to translate free-form natural language feedback and prior kn

Yehonathan Refael, Amit Aides, Aviad Barzilai, George Leifman

Open-vocabulary object detection (OVD) models offer remarkable flexibility by detecting objects from arbitrary text queries. However, their zero-shot performance in specialized domains like Remote Sensing (RS) is often compromised by the inherent ambiguity of natural language, limiting critical downstream applications. For instance, an OVD model may struggle

Bartosz Bieganowski, Pietro d'Avenia, Jacopo Schino, Daniel Strzelecki

In the paper, we prove the existence of a positive and essentially bounded solution to a Lichnerowicz equation in the Einstein-scalar field theory on a closed manifold with non-constant mean curvature. In particular, the non-constant mean curvature gives rise to supercritical terms in the equation, on top of singular ones. We employ a recent fixed-point argu

Haoyu Huang, Hong Ting Tsang, Jiaxin Bai, Xi Peng

Retrieval-augmented generation (RAG) has shown some success in augmenting large language models (LLMs) with external knowledge. However, as a non-parametric knowledge integration paradigm for LLMs, RAG methods heavily rely on external retrieval modules and the retrieved textual context prior. Especially for very large scale knowledge augmentation, they would

Soumyaditya Das, Soumyajyoti Biswas, Muktish Acharyya, Bikas K. Chakrabarti

We briefly review the early development of the mean-field dynamics for cooperatively interacting quantum many-body systems, mapped to pseudo-spin (Ising-like) systems. We start with (Anderson, 1958) pseudo-spin mapping of the BCS (1957) Hamiltonian of superconductivity, reducing it to a mean-field Hamiltonian of XY (or effectively Ising) model in a transvers

Graciana Puentes

A combination of statistical inference and machine learning (ML) schemes has been utilized to create a thorough understanding of coarse experimental data based on Zernike variables characterizing optical aberrations in fluidic lenses. A classification of surplus-response variables through tolerance manipulation was included to unravel the dimensional aspect

Matthew Chaffe, Gabriele Rembado, Daisuke Yamakawa

The wild de Rham spaces parameterize isomorphism classes of (stable) meromorphic connections, defined on principal bundles over wild Riemann surfaces. Working on the Riemann sphere, we will deformation-quantize the standard open part of de Rham spaces, which corresponds to the moduli of linear ordinary differential equations with meromorphic coefficients. We

Ling Liu, Jun Tian, Li Yi

4D panoptic segmentation in a streaming setting is critical for highly dynamic environments, such as evacuating dense crowds and autonomous driving in complex scenarios, where real-time, fine-grained perception within a constrained time budget is essential. In this paper, we introduce 4DSegStreamer, a novel framework that employs a Dual-Thread System to effi

Faqiang Yuan, Haida Li, Shengzhi Li, Yongge Ma

The degree of freedom of the scalar field in scalar-tensor gravity is employed as "time" to deparametrize the Hamiltonian constraint of the theory. The deparametrized system is then nonperturbatively quantized by the approach of loop quantum gravity. This results in a discrete time evolution of the physical states with respect to the gravitational degree of

Massa Baali, Rita Singh, Bhiksha Raj

Self-supervised speech models have achieved remarkable success on content-driven tasks, yet they remain limited in capturing speaker-discriminative features critical for verification, diarization, and profiling applications. We introduce \textsc{DELULU}, a speaker-aware self-trained foundational model that addresses this limitation by incorporating speaker-i

Vaishnavi Visweswaraiah, Tanvi Banerjee, William Romine

Suicide prediction is the key for prevention, but real data with sufficient positive samples is rare and causes extreme class imbalance. We utilized machine learning (ML) to build the model and deep learning (DL) techniques, like Generative Adversarial Networks (GAN), to generate synthetic data samples to enhance the dataset. The initial dataset contained 65