March 2025 arXiv papers — page 9

Yijie Zheng, Bangjun Xiao, Lei Shi, Xiaoyang Li

Multimodal large language models (MLLMs), such as GPT-4o, are garnering significant attention. During the exploration of MLLM training, we identified Modality Composition Incoherence, a phenomenon that the proportion of a certain modality varies dramatically across different examples. It exacerbates the challenges of addressing mini-batch imbalances, which l

Yi Su, Dian Yu, Linfeng Song, Juntao Li

Reinforcement learning with verifiable rewards (RLVR) has demonstrated significant success in enhancing mathematical reasoning and coding performance of large language models (LLMs), especially when structured reference answers are accessible for verification. However, its extension to broader, less structured domains remains unexplored. In this work, we inv

Oscar Groppfeldt, Joel Davidsson, Rickard Armiento

Point defects in semiconductors offer a promising platform for advancing quantum technologies due to their localized energy states and controllable spin properties. Prior research has focused on a limited set of defects within materials such as diamond, silicon carbide, and hexagonal boron nitride. We present a high-throughput study to systematically identif

Phat Lam, Lam Pham, Dat Tran, Alexander Schindler

In this paper, we present an audio analyzer assistant tool designed for a wide range of audio-based surveillance applications (This work is a part of our DEFAME FAKES and EUCINF projects). The proposed tool, refered to as Aud-Sur, comprises two main phases Audio Analysis and Audio Retrieval, respectively. In the first phase, multiple open-source audio models

Shaull Almagor, Guy Arbel, Sarai Sheinvald

We show that the determinization problem for min-plus (tropical) weighted automata is decidable, thus resolving this long-standing open problem. In doing so, we develop a new toolbox for analyzing and reasoning about the run-structure of nondeterministic automata.

Saikat Gayen, Jais Kumar, Prasun Dutta, Khandakar Md Asif Elahi

Observation of multifrequency angular power spectrum of the redshifted 21-cm brightness temperature fluctuation from the neutral hydrogen holds the key to understand the structure formation and its evolution during the reionization and post-reionization era. A major challenge in observing the neutral hydrogen arises from presence of strong foreground signals

Emmanouil Georgios Lionis, Jia-Huei Ju

Document retrieval is one of the most challenging tasks in Information Retrieval. It requires handling longer contexts, often resulting in higher query latency and increased computational overhead. Recently, Learned Sparse Retrieval (LSR) has emerged as a promising approach to address these challenges. Some have proposed adapting the LSR approach to longer d

Farhana Javed, Engin Zeydan, Josep Mangues-Bafalluy, Kapal Dev

As edge computing gains prominence in Internet of Things (IoTs), smart cities, and autonomous systems, the demand for real-time machine intelligence with low latency and model reliability continues to grow. Federated Learning (FL) addresses these needs by enabling distributed model training without centralizing user data, yet it remains reliant on centralize

A. Ali, M. Barone, D. Britzger, A. Cooper-Sarkar

The European Strategy for Particle Physics (ESPP) - 2026 update is taking place in a turbulent international climate. Many of the norms that have governed relations between states for decades are being broken or challenged. The future progress of science in general, and particle physics in particular, will depend on our ability to maintain peaceful internati

Yahya Aalaila, Gerrit Großmann, Sumantrak Mukherjee, Jonas Wahl

Counterfactual reasoning, a cornerstone of human cognition and decision-making, is often seen as the 'holy grail' of causal learning, with applications ranging from interpreting machine learning models to promoting algorithmic fairness. While counterfactual reasoning has been extensively studied in contexts where the underlying causal model is well-defined,

Ilir Tahiraj, Jeremialie Swadiryus, Felix Fent, Markus Lienkamp

The goal of extrinsic calibration is the alignment of sensor data to ensure an accurate representation of the surroundings and enable sensor fusion applications. From a safety perspective, sensor calibration is a key enabler of autonomous driving. In the current state of the art, a trend from target-based offline calibration towards targetless online calibra

Ilir Tahiraj, Markus Edinger, Dominik Kulmer, Markus Lienkamp

In autonomous systems, sensor calibration is essential for safe and efficient navigation in dynamic environments. Accurate calibration is a prerequisite for reliable perception and planning tasks such as object detection and obstacle avoidance. Many existing LiDAR calibration methods require overlapping fields of view, while others use external sensing devic

Swarnava Bhattacharyya, Umapada Pal, Tapabrata Chakraborti

Deep learning based diagnostic AI systems based on medical images are starting to provide similar performance as human experts. However these data hungry complex systems are inherently black boxes and therefore slow to be adopted for high risk applications like healthcare. This problem of lack of transparency is exacerbated in the case of recent large founda

Leonardo Massai, Muhammad Zakwan, Giancarlo Ferrari-Trecate

Structured state-space models (SSMs) have recently emerged as a powerful architecture at the intersection of machine learning and control, featuring layers composed of discrete-time linear time-invariant (LTI) systems followed by pointwise nonlinearities. These models combine the expressiveness of deep neural networks with the interpretability and inductive

Tatsuya Kubo, Daichi Tokuda, Tomoya Nagatani, Masayuki Usui

General matrix-vector multiplication (GeMV) remains a critical latency bottleneck in large language model (LLM) inference, even with quantized low-bit models. Processing-Using-DRAM (PUD), an analog in-DRAM computing technique, has the potential to repurpose on-device DRAM as a GeMV engine, offering additional high-throughput processing capabilities to widesp

K. A. Muthukumar, Dhruva Nandi, Priya Ranjan, Krithika Ramachandran

Cardiovascular diseases (CVD) are a predominant health concern globally, emphasizing the need for advanced diagnostic techniques. In our research, we present an avant-garde methodology that synergistically integrates ECG readings and retinal fundus images to facilitate the early disease tagging as well as triaging of the CVDs in the order of disease priority

Jean-François Delmas, Dylan Dronnier, Pierre-André Zitt

This present results lay the foundations for the study of the optimal allocation of vaccine in the simple epidemiological SIS model where one consider a very general heterogeneous population. In the present setting each individual has a type x belonging to a general space, and a vaccination strategy is a function $\eta$ where $\eta$(x) $\in$ [0, 1] represent

Keyu Wang, Peihao Huang

We study an interacting two-body model with adjustable spin tunneling in the context of the double SSH chains for a quantum dot system. We discovered that varying interaction strengths and spin tunneling significantly influence the properties of correlated edge states in the energy spectrum obtained through exact diagonalization. We observe that stronger int

Saroj Prasad Chhatoi, Jean B Lasserre

We consider the LP in standard form min {c T x\,: Ax = b; x $\ge$ 0} and inspired by $\epsilon$-regularization in Optimal Transport, we introduce its $\epsilon$-regularization ''min {c T x + $\epsilon$ f (x)\,: Ax = b; x $\ge$ 0}'' via the (convex) Boltzmann-Shannon entropy f (x)\,:= i x i ln x i . We also provide a similar regularization for the semidefinit

Katsuyuki Hagiwara

In-context learning is a remarkable property of transformers and has been the focus of recent research. An attention mechanism is a key component in transformers, in which an attention matrix encodes relationships between words in a sentence and is used as weights for words in a sentence. This mechanism is effective for capturing language representations. Ho

Hongbin Song

This paper presents the findings of pedagogical research on the efficacy of a virtual laboratory platform in general education courses on quantum information science. Specifically, a virtual laboratory activity based on the Bell test has been developed using a commercially available Quantum Optical Simulation Laboratory, QLab. The experiential activity is de

Youssef Abdulghani, Anne M Lohfink, Jaiverdhan Chauhan

Investigations of the Galactic black hole low-mass X-ray binaries (BH-LMXBs) offer valuable insights into the elusive black hole population in the Milky Way. Motivated by recent tensions in the natal kick velocity distribution and BH mass distribution of BH-LMXBs, we revisit the spatial distribution of the Galactic BH-LMXBs using a new set of distance measur

Chris Brogly, Connor McElroy

The release of advanced Large Language Models (LLMs) such as ChatGPT and Copilot is changing the way text is created and may influence the content that we find on the web. This study investigated whether the release of these two popular LLMs coincided with a change in writing style in headlines and links on worldwide news websites. 175 NLP features were obta

Ilayda Yaman, Guoda Tian, Dino Pjanic, Fredrik Tufvesson

Radio-based localization in dynamic environments, such as urban and vehicular settings, requires systems that efficiently adapt to varying signal conditions and environmental changes. Factors like multipath interference and obstructions introduce different levels of complexity that affect the accuracy of the localization. Although generalized models offer br

E. Bonnet, B. Borderie, R. Bougault, A. Chbihi

In this article, we study the production of Hydrogen and Helium isotopes in heavy-ion collisions in the incident energy range between 80 and 150 MeV/nucleon. We compare their inclusive multiplicities emitted in the transverse plane of the reaction with the predictions given by the thermal model. As a first step, we validate the choice of this approach to des

Dario Garcia-Gasulla, Gokcen Kestor, Emanuele Parisi, Miquel Albertí-Binimelis

The rapid advancements in LLMs have driven the adoption of generative AI in various domains, including Electronic Design Automation (EDA). Unlike traditional software development, EDA presents unique challenges, as generated RTL code must not only be syntactically correct and functionally accurate but also synthesizable by hardware generators while meeting p

Robin Guillard, Vincent Philippe, Adam Hessas, Brice Faraut

Background: Tinnitus, defined as the conscious awareness of a noise without any identifiable corresponding external acoustic source, can be modulated by various factors. Among these factors, tinnitus patients commonly report drastic increases of tinnitus loudness following nap sleep. Previous studies have suggested that this clinical pattern could be attribu

Toranosuke Matsubara, Kazuki Yamamoto, Akihisa Koga

We study the dynamics under continuous measurements for free fermions in a quasiperiodic potential by using the Aubry-Andr\'{e}-Harper model with hopping rate $J$ and potential strength $V$. On the basis of the quantum trajectory method, we obtain the phase diagram for the steady-state entanglement entropy and demonstrate that robust logarithmic system-size

Xiaoqing Guo, Wuyang Li, Yixuan Yuan

Generalized zero-shot semantic segmentation (GZS3) aims to achieve the human-level capability of segmenting not only seen classes but also novel class regions unseen in the training data through introducing the bridge of semantic representations, e.g., word vector. While effective, the way of utilizing one semantic representation to associate the correspondi

Davide Tebaldi, Roberto Zanasi

A powerful tool in control and systems engineering is represented by Nyquist plots, for which a qualitative representation often gives a clearer visualization of the frequency response function that is typically not given by computer programs, especially if portions of the Nyquist plot extend to infinity. This letter addresses the graphical analysis of the f

Shiyi Yang, Zhibo Hu, Xinshu Li, Chen Wang

Large language model (LLM)-powered agents are increasingly used in recommender systems (RSs) to achieve personalized behavior modeling, where the memory mechanism plays a pivotal role in enabling the agents to autonomously explore, learn and self-evolve from real-world interactions. However, this very mechanism, serving as a contextual repository, inherently

Yingwei Ma, Yongbin Li, Yihong Dong, Xue Jiang

Recent advancements in software engineering agents have demonstrated promising capabilities in automating program improvements. However, their reliance on closed-source or resource-intensive models introduces significant deployment challenges in private environments, prompting a critical question: \textit{How can personally deployable open-source LLMs achiev

Chris Brogly, Saif Rjaibi, Charlotte Liang, Erica Lam

Small Language Models (SLMs) have potential to be used for automatically labelling and identifying aspects of text data for medicine/health-related purposes from documents and the web. As their resource requirements are significantly lower than Large Language Models (LLMs), these can be deployed potentially on more types of devices. SLMs often are benchmarke

I. Dhivviyanandam, A. Lourdusamy, S. Kither Iammal, K. Christy Rani

Consider a configuration of pebbles on the vertices of a connected graph. A pebbling move is to remove two pebbles from a vertex and to place one pebble at the neighbouring vertex of the vertex from which the pebbles are removed. For a positive integer $t$, with every configuration of $\pi_t(G)$(least positive integer) pebbles, if we can transfer $t$ pebbles

Marie-Christine Volk, Anne Sergent, Didier Lucor, Michael Mommert

We present a method to infer temperature fields from stereo particle-image velocimetry (PIV) data in turbulent Rayleigh-B\'enard convection (RBC) using Physics-informed neural networks (PINNs). The physical setup is a cubic RBC cell with Rayleigh number $\text{Ra}=10^7$ and Prandtl number $\text{Pr}=0.7$. With data only available in a vertical plane $A:x=x_0

Xingyu Yang, Chantal Hareau, Tristan da Câmara Santa Clara Gomes, Jack Gartside

Skyrmions are topological structures defined by a winding vector configuration that yields a quantized topological charge. In magnetic materials, skyrmions manifest as stable, mobile spin textures, positioning them at the forefront of spintronics research. Meanwhile, their optical counterparts unlock new possibilities for manipulating and directing light at

Shuang Miao, Shiwu Yang, Pin Yu

This paper is devoted to presenting a rigorous mathematical derivation for the classical phenomenon in Maxwell's theory that a charged particle moves along a straight line in a constant electromagnetic field if the initial velocity is parallel to the constant electromagnetic field. The particle is modeled by scaled solitons to a class of nonlinear Klein-Gord

Xuan Luo, Weizhi Wang, Xifeng Yan

Various layer-skipping methods have been proposed to accelerate token generation in large language models (LLMs). However, limited attention has been paid to a fundamental question: How do computational demands vary across the generation of different tokens? In this work, we introduce FlexiDepth, a method that dynamically adjusts the number of Transformer la

Anita Graser

Crowd and flow predictions have been extensively studied in mobility data science. Traditional forecasting methods have relied on statistical models such as ARIMA, later supplemented by deep learning approaches like ST-ResNet. More recently, foundation models for time series forecasting, such as TimeGPT, Chronos, and LagLlama, have emerged. A key advantage o

Fabian Walter, Eduardo Banados, Chris Carilli, Marcel Neeleman

We present high-angular resolution (0.068", ~400pc) ALMA imaging of the [CII] line and dust continuum emission of PSO J352.4034-15.3373, a radio-loud quasar at z=5.83. The observations reveal a remarkably close match between the orientation of the [CII] and thermal dust emission mapped by ALMA, and radio synchrotron emission of a radio jet previously mapped

Bosung Kim, Kyuhwan Lee, Isu Jeong, Jungmin Cheon

We present On-device Sora, the first model training-free solution for diffusion-based on-device text-to-video generation that operates efficiently on smartphone-grade devices. To address the challenges of diffusion-based text-to-video generation on computation- and memory-limited mobile devices, the proposed On-device Sora applies three novel techniques to p

Benjamin Bogenberger, Johannes Bürger, Vladislav Nenchev

Self-driving vehicles rely on sensory input to monitor their surroundings and continuously adapt to the most likely future road course. Predictive trajectory planning is based on snapshots of the (uncertain) road course as a key input. Under noisy perception data, estimates of the road course can vary significantly, leading to indecisive and erratic steering

Fabian L. Thiemann, Thiago Reschützegger, Massimiliano Esposito, Tseden Taddese

Molecular dynamics (MD) simulations play a crucial role in scientific research. Yet their computational cost often limits the timescales and system sizes that can be explored. Most data-driven efforts have been focused on reducing the computational cost of accurate interatomic forces required for solving the equations of motion. Despite their success, howeve

Zhongnan Cai, Yingying Wang, Hui Zheng, Panwang Pan

Recently, deep learning-based pan-sharpening algorithms have achieved notable advancements over traditional methods. However, deep learning-based methods incur substantial computational overhead during inference, especially with large images. This excessive computational demand limits the applicability of these methods in real-world scenarios, particularly i

Takeshi Fukasawa

This study empirically investigates firms' incentives on the choice of product durability, and its social optimality, by developing a dynamic structural model of durable goods with forward-looking consumers and oligopolistic multi-product firms. Based on the observations of the light bulb market, it specifies a model where firms produce multiple products wit

Yuchen Liu, Junhao Hu, Yingdi Shan, Ge Li

Rewriting C code in Rust provides stronger memory safety, yet migrating large codebases such as the 32-million-line Linux kernel remains challenging. While rule-based translators (e.g., C2Rust) provide accurate yet largely unsafe Rust programs, recent Large Language Model (LLM) approaches produce more idiomatic, safe Rust programs but frequently exhibit "laz

Lorenzo Barban, Gianluca Occhetta, Luis E. Sol á Conde

We construct geometric realizations -- projective algebraic versions of cobordisms -- for birational maps between Mori Dream Spaces. We show that these geometric realizations are Mori Dream Spaces, as well, and that they can be constructed so that they induce factorizations of the original birational maps as compositions of wall-crossings. In the case of tor

Adrian Lopez-Rosales, Borja Ferreiro, Jose Andrade, Andreas Kerstan

Microplastic pollution studies depend on reliable identification of the suspicious particles. Out of the various analytical techniques available to characterize them, infrared transflectance using a tuneable mid-IR quantum cascade laser is a high-throughput state-of-the-art imaging option, specifically Agilent QCL-LDIR (Quantum Cascade Laser Direct Infrared

Frederic V. Hessman, Andrew Collier Cameron, Keith Horne

We report the detection of whisky in the atmosphere of the extrasolar super-Earth planet GJ 1132b from transmission spectroscopic data. It is seen both in atmospheric absorption as well as in chromospheric emission, the latter probably due to the intense heating of the co-rotating planet's day-side surface. This detection cannot be explained using natural so

Haoran Shen, Peixian Zhuang, Jiahao Kou, Yuxin Zeng

Segment Anything Models (SAMs), as vision foundation models, have demonstrated remarkable performance across various image analysis tasks. Despite their strong generalization capabilities, SAMs encounter challenges in fine-grained detail segmentation for high-resolution class-independent segmentation (HRCS), due to the limitations in the direct processing of

Nikhil Bartake, See Toh Zi Jie, Carmen Wong Jiawen, Michael Kasper

This paper introduces ObfusQate, a novel tool that conducts obfuscations using quantum primitives to enhance the security of both classical and quantum programs. We have designed and implemented two primary categories of obfuscations: quantum circuit level obfuscation and code level obfuscation, encompassing a total of eight distinct methods. Quantum circuit

Adrian Lopez-Rosales, Jose Andrade, Borja Ferreiro, Soledad Muniategui

Microplastics (MPs) are ubiquitous in all ecosystems, affecting wildlife and, ultimately, human health. The complexity of natural samples plus the unspecificity of their treatments to isolate polymers renders the characterization of thousands of particles impractical for environmental monitoring using conventional spectroscopic techniques. Two primary soluti

Ze Xu, Raphael Saiseau, Olinka Ramírez Soto, Stefan Karpitschka

Wetting of micropatterned surfaces is ubiquitous in nature and key to many technological applications like spray cooling, inkjet printing, and semiconductor processing. Overcoming the intrinsic, chemistry- and topography-governed wetting behaviors often requires specific materials which limits applicability. Here, we show that spreading and wicking of water

Tianqi Chen, Wei Huang, Qiang Wu, Li Yang

The traditional method for designing branch-line couplers involves a trial-and-error optimization process that requires multiple design iterations through electromagnetic (EM) simulations. Thus, it is extremely time consuming and labor intensive. In this paper, a novel machine-learning-based framework is proposed to tackle this issue. It integrates artificia

Ahmed Zaoui, Clément Dombry

Selective prediction, where a model has the option to abstain from making a decision, is crucial for machine learning applications in which mistakes are costly. In this work, we focus on distributional regression and introduce a framework that enables the model to abstain from estimation in situations of high uncertainty. We refer to this approach as distrib

Jinwei Su, Yinghui Xia, Yiqun Duan, Jun Du

Large language models (LLMs) have demonstrated strong potential and impressive performance in automating the generation and optimization of workflows. However, existing approaches are marked by limited reasoning capabilities, high computational demands, and significant resource requirements. To address these issues, we propose DebFlow, a framework that emplo

Madoka Horie, Takuya Yamauchi

In this paper, we explicitly compute two kinds of algebraic Belyi functions on Bring's curve. One is related to a congruence subgroup of ${\rm SL}_2(\mathbb{Z})$ and the other is related to a congruence subgroup of the triangle group $\Delta(2,4,5)\subset \SL_2(\R)$. To carry out the computation, we use elliptic cusp forms of weight 2 for the former case and

Ine Gevers, Victor De Marez, Luna De Bruyne, Walter Daelemans

In this study, we take a closer look at how Winograd schema challenges can be used to evaluate common sense reasoning in LLMs. Specifically, we evaluate generative models of different sizes on the popular WinoGrande benchmark. We release WinoWhat, a new corpus, in which each instance of the WinoGrande validation set is paraphrased. Additionally, we evaluate

Man Shen, Taiki Inoue, Mengyue Wang, Yuanjia Liu

Defects in single-walled carbon nanotubes (SWCNTs) degrade their mechanical,electrical, and thermal properties, limiting their potential applications. To realize the diverse applications of SWCNTs, it is essential to enhance their crystallinity through effective defect healing. However, traditional thermal treatments typically require temperatures above 1800

Jiangnan Li, Thuy-Trang Vu, Christian Herold, Amirhossein Tebbifakhr

Naive joint training of large language models (LLMs) for multilingual preference alignment can suffer from negative interference. This is a known issue in multilingual training, where conflicting objectives degrade overall performance. However, the impact of this phenomenon in the context of multilingual preference alignment remains largely underexplored. To

Rong Kang, Shuai Wang, Tieying Zhang, Xianghong Xu

Virtual index, also known as hypothetical indexes, play a crucial role in database query optimization. However, with the rapid advancement of cloud computing and AI-driven models for database optimization, traditional virtual index approaches face significant challenges. Cloud-native environments often prohibit direct conducting query optimization process on

Lucas Heublein, Nisha L. Raichur, Tobias Feigl, Tobias Brieger

The accuracy and reliability of vehicle localization on roads are crucial for applications such as self-driving cars, toll systems, and digital tachographs. To achieve accurate positioning, vehicles typically use global navigation satellite system (GNSS) receivers to validate their absolute positions. However, GNSS-based positioning can be compromised by int

Liat Lavi

This short paper puts forward a vision for a new democratic model enabled by the recent technological advances in agentic AI. It therefore opens with drawing a clear and concise picture of the model, and only later addresses related proposals and research directions, and concerns regarding feasibility and safety. It ends with a note on the timeliness of this

Chaopeng Luo, Yuanliang Zhang, Haochen He, Zhouyang Jia

Runtime misconfiguration can lead to software performance degradation and even cause failure. Developers typically perform sanity checks during the configuration parsing stage to prevent invalid parameter values. However, we discovered that even valid values that pass these checks can also lead to unexpected severe consequences. Our study reveals the underly

Ear Philippe, Di Bernardino Elena, Laloë Thomas, Troin Magali

Highly resoluted and accurate daily precipitation data are required for impact models to perform adequately and to correctly measure high-risk events' impact. In order to produce such data, bias-correction is often needed. Most of those statistical methods correct the probability distributions of daily precipitation by modeling them using either empirical or

Yuxiao Sun, Yao Zhao, Meiqin Liu, Chao Yao

Nowadays, more and more video transmissions primarily aim at downstream machine vision tasks rather than humans. While widely deployed Human Visual System (HVS) oriented video coding standards like H.265/HEVC and H.264/AVC are efficient, they are not the optimal approaches for Video Coding for Machines (VCM) scenarios, leading to unnecessary bitrate expendit

Fengxiang Wang, Hongzhen Wang, Mingshuo Chen, Di Wang

The astonishing breakthrough of multimodal large language models (MLLMs) has necessitated new benchmarks to quantitatively assess their capabilities, reveal their limitations, and indicate future research directions. However, this is challenging in the context of remote sensing (RS), since the imagery features ultra-high resolution that incorporates extremel

Semyon Yakubovich

An index transform, involving the square of Whittaker's function is introduced and investigated. The corresponding inversion formula is established. Particular cases cover index transforms of the Lebedev type with products of the modified Bessel functions.

Nobuya Mori, Hajime Tanaka, Jo Okada

The motion of electrons under homogeneously applied electric fields in low-dimensional systems with non-zero off-diagonal effective mass (ODEM) is studied. The equation describing the time evolution of a probability coefficient of finding an electron in a subband is derived using the Krieger-Iafrate theory in the effective mass approximation. It is shown tha

Xian-Xian Liu, Yuanyuan Wei, Mingkun Xu, Yongze Guo

Early detection of gastric cancer, a leading cause of cancer-related mortality worldwide, remains hampered by the limitations of current diagnostic technologies, leading to high rates of misdiagnosis and missed diagnoses. To address these challenges, we propose an integrated system that synergizes advanced hardware and software technologies to balance speed-

Zhecheng Li, Guoxian Song, Yujun Cai, Zhen Xiong

Modern Vision-Language Models (VLMs) exhibit remarkable visual and linguistic capabilities, achieving impressive performance in various tasks such as image recognition and object localization. However, their effectiveness in fine-grained tasks remains an open question. In everyday scenarios, individuals encountering design materials, such as magazines, typog

Linghao Feng, Dongcheng Zhao, Sicheng Shen, Yi Zeng

Lateral connection is a fundamental feature of biological neural circuits, facilitating local information processing and adaptive learning. In this work, we integrate lateral connections with a substructure selection network to develop a novel diffusion model based on spiking neural networks (SNNs). Unlike conventional artificial neural networks, SNNs employ

Jiangjie Qiu, Hou Hei Lam, Xiuyuan Hu, Wentao Li

Organic photovoltaic (OPV) materials offer a promising avenue toward cost-effective solar energy utilization. However, optimizing donor-acceptor (D-A) combinations to achieve high power conversion efficiency (PCE) remains a significant challenge. In this work, we propose a framework that integrates large-scale pretraining of graph neural networks (GNNs) with

Yun Li, Yiming Zhang, Tao Lin, Xiangrui Liu

The use of Multimodal Large Language Models (MLLMs) as an end-to-end solution for Embodied AI and Autonomous Driving has become a prevailing trend. While MLLMs have been extensively studied for visual semantic understanding tasks, their ability to perform precise and quantitative spatial-temporal understanding in real-world applications remains largely unexa

Md Mahfuz Al Hasan, Mahdi Zaman, Abdul Jawad, Alberto Santamaria-Pang

Transformer-based architectures have advanced medical image analysis by effectively modeling long-range dependencies, yet they often struggle in 3D settings due to substantial memory overhead and insufficient capture of fine-grained local features. We address these limitations with WaveFormer, a novel 3D-transformer that: i) leverages the fundamental frequen

Jun Honda

We examine how career concerns influence the behavior and mobility of financial advisers. Drawing on a uniquely comprehensive matched panel that combines employer-employee data with a longstanding national ranking, our study tests predictions from classic career concerns models and tournament theory. Our analysis shows that, in the early stages of their care

Yuanyuan Wang, Hangting Chen, Dongchao Yang, Weiqin Li

We propose Universal target audio Separation (UniSep), addressing the separation task on arbitrary mixtures of different types of audio. Distinguished from previous studies, UniSep is performed on unlimited source domains and unlimited source numbers. We formulate the separation task as a sequence-to-sequence problem, and a large language model (LLM) is used

Suresh Govindarajan, Jagannath Santara

The holomorphic bootstrap attempts to classify rational conformal field theories. The straight ahead approach is hard to implement when the number of characters become large. We combine all characters of an RCFT to form a vector valued modular form with multiplier. Using known results from the theory of vector valued modular forms, given a known RCFT, we obt

Eunseok Hwang, Myung-Ki Cheoun, Dukjae Jang

We explore the averaged fusion reactivity of the $p+{^{11}{\rm B}}$ reaction in tabletop laser experiments using a plasma expansion model. We investigate the energy distribution of proton beams accelerated by lasers as a function of electron temperature $T_e$ and the dimensionless acceleration time $\omega_{pi} t_{\rm acc}$, where $\omega_{pi}$ is the ion pl

Manuel Scheibl, Birte Richter, Alissa Müller, Michael Beetz

This research addresses the question, which characteristics a cognitive architecture must have to leverage the benefits of natural language in Co-Constructive Task Learning (CCTL). To provide context, we first discuss Interactive Task Learning (ITL), the mechanisms of the human memory system, and the significance of natural language and multi-modality. Next,

Rajat De, Dominik Kempa

Word Break is a prototypical factorization problem in string processing: Given a word $w$ of length $N$ and a dictionary $\mathcal{D} = \{d_1, d_2, \ldots, d_{K}\}$ of $K$ strings, determine whether we can partition $w$ into words from $\mathcal{D}$. We propose the first algorithm that solves the Word Break problem over the SLP-compressed input text $w$. Spe

Weiguo Yin

The one-dimensional (1D) $J_1$-$J_2$ $q$-state Potts model is solved exactly for arbitrary $q$ by analytically block-diagonalizing the original $q^2\times q^2$ transfer matrix into a simple $2\times 2$ maximally symmetric subspace, based on using OpenAI's reasoning model o3-mini-high to exactly solve the $q=3$ case. Furthermore, by matching relevant subspace

ZiXin Lin, Nur Fariha Syaqina Zulkepli

Topological Data Analysis (TDA) has emerged as a powerful tool for extracting meaningful features from complex data structures, driving significant advancements in fields such as neuroscience, biology, machine learning, and financial modeling. Despite its success, the integration of TDA with time-series prediction remains underexplored due to three primary c

Jinwei Gao

We investigate the space of Hermitian metrics on a fixed complex vector bundle. This infinite-dimensional space has appeared in the study of Hermitian-Einstein structures, where a special L2-type Riemannian metric is introduced. We compute the metric spray, geodesics and curvature associated to this metric, and show that the exponential map is a diffeomorphs

Xudong Wang, Xiuqi Zhang, Bowen Chen, Yifan Zhu

Hybrid integrated quantum photonics combines solid-state artificial atoms with reconfigurable photonic circuits, enabling scalable chip-based quantum networks. Self-assembled quantum dots (QDs) are ideal for this goal due to their ability to generate highly indistinguishable single photons with exceptional brightness. Integrating QDs into low-loss photonic c

Nitin Tomar

For $ 0 < r < 1 $, let $ \mathbb{A}_r = \{ z \in \mathbb{C} : r < |z| < 1 \} $ be the annulus with boundary $ \partial \overline{\mathbb{A}}_r = \mathbb{T} \cup r\mathbb{T} $, where $ \mathbb{T} $ is the unit circle in the complex plane $\mathbb C$. We study the class of operators \[ C_{1,r} = \{ T : T \text{ is invertible and } \|T\|, \|rT^{-1}\| \leq 1 \},

Kai Xie, Attila Szolnoki

Understanding and resolving cooperation dilemmas are key challenges in evolutionary game theory, which have revealed several mechanisms to address them. This paper investigates the comprehensive influence of multiple reputation-related components on public cooperation. In particular, cooperative investments in public goods game are not fixed but simultaneous

Jin Zhou, Yi Zhou, Hongliang Yang, Pengfei Xu

In the field of sketch generation, raster-format trained models often produce non-stroke artifacts, while vector-format trained models typically lack a holistic understanding of sketches, leading to compromised recognizability. Moreover, existing methods struggle to extract common features from similar elements (e.g., eyes of animals) appearing at varying po

Yu Zhou, Dian Zheng, Qijie Mo, Renjie Lu

In this work, we present DEcoupLEd Distillation To Erase (DELETE), a general and strong unlearning method for any class-centric tasks. To derive this, we first propose a theoretical framework to analyze the general form of unlearning loss and decompose it into forgetting and retention terms. Through the theoretical framework, we point out that a class of pre

Masatoshi Kitagawa

Vinberg--Kimel'fel'd [Funct. Anal. Appl., 1978] established that a quasi-projective normal $G$-variety $X$ is spherical if and only if $G$-modules on the spaces $\Gamma(X, \mathcal{L})$ of global sections of $G$-equivariant line bundles are multiplicity-free. This result was generalized by Kobayashi--Oshima [Adv. Math., 2013] and several researchers to (dege

Yujin Huang, Zhi Zhang, Qingchuan Zhao, Xingliang Yuan

On-device deep learning (DL) has rapidly gained adoption in mobile apps, offering the benefits of offline model inference and user privacy preservation over cloud-based approaches. However, it inevitably stores models on user devices, introducing new vulnerabilities, particularly model-stealing attacks and intellectual property infringement. While system-lev

Jingyi Zhou, Peng Ye, Haoyu Zhang, Jiakang Yuan

Iterative-based methods have become mainstream in stereo matching due to their high performance. However, these methods heavily rely on labeled data and face challenges with unlabeled real-world data. To this end, we propose a consistency-aware self-training framework for iterative-based stereo matching for the first time, leveraging real-world unlabeled dat

Dizhan Xue, Shengsheng Qian, Chuanrui Hu, Changsheng Xu

Short-video platforms have gained immense popularity, captivating the interest of millions, if not billions, of users globally. Recently, researchers have highlighted the significance of analyzing the propagation of short-videos, which typically involves discovering commercial values, public opinions, user behaviors, etc. This paper proposes a new Short-vide

Fabian Göttsch, Shuangyang Li, Lorenzo Miretti, Giuseppe Caire

In this paper, we perform a comparative study of common wireless communication waveforms, namely the single carrier (SC), orthogonal frequency-division multiplexing (OFDM), and orthogonal time-frequency-space (OTFS) modulation in a millimeter wave (mmWave) downlink multi-connectivity scenario, where multiple access points (APs) jointly serve a given user und

DUNE Collaboration, A. Abed Abud, R. Acciarri, M. A. Acero

The Proton Improvement Plan (PIP-II) to the FNAL accelerator chain and the Long-Baseline Neutrino Facility (LBNF) will provide the world's most intense neutrino beam to the Deep Underground Neutrino Experiment (DUNE) enabling a wide-ranging physics program. This document outlines the significant contributions made by European national laboratories and instit

DUNE Collaboration, A. Abed Abud, R. Acciarri, M. A. Acero

The international collaboration designing and constructing the Deep Underground Neutrino Experiment (DUNE) at the Long-Baseline Neutrino Facility (LBNF) has developed a two-phase strategy toward the implementation of this leading-edge, large-scale science project. The ambitious physics program of Phase I and Phase II of DUNE is dependent upon deployment and

Minhyuk Jang, Astghik Hakobyan, Insoon Yang

State estimation in the presence of uncertain or data-driven noise distributions remains a critical challenge in control and robotics. Although the Kalman filter is the most popular choice, its performance degrades significantly when distributional mismatches occur, potentially leading to instability or divergence. To address this limitation, we introduce a

Bin Shen

In this manuscript, we extend the global gradient estimates for positive solutions to the heat equation under a general compact Finsler $CD(-K,N)$ geometric flow and derive the corresponding Harnack inequality.

Shuta Funakoshi, Yuichi Koga, Hajime Otsuka

We study modular symmetries in non-supersymmetric heterotic string theories on toroidal backgrounds with Wilson line modulus, constructed by stringy Scherk-Schwartz compactification. In particular, we focus on a subgroup of the T-duality group $O(D+16,D,\mathbb{Z})$ with $D=2$ given by an outer automorphism of the Narain lattice, which can be mapped to the S

Lu Fan, Jiashu Pu, Rongsheng Zhang, Xiao-Ming Wu

Task-oriented Dialogue Systems (TODS) often face the challenge of encountering new intents. New Intent Discovery (NID) is a crucial task that aims to identify these novel intents while maintaining the capability to recognize existing ones. Previous efforts to adapt TODS to new intents have struggled with inadequate semantic representation or have depended on