May 2025 arXiv papers — page 101

Katharina Winter, Abhishek Vivekanandan, Rupert Polley, Yinzhe Shen

Generative AI (GenAI) is rapidly advancing the field of Autonomous Driving (AD), extending beyond traditional applications in text, image, and video generation. We explore how generative models can enhance automotive tasks, such as static map creation, dynamic scenario generation, trajectory forecasting, and vehicle motion planning. By examining multiple gen

Mathew Vanherreweghe, Lirandë Pira, Patrick Rebentrost

We introduce cumulative polynomial Kolmogorov-Arnold networks (CP-KAN), a neural architecture combining Chebyshev polynomial basis functions and quadratic unconstrained binary optimization (QUBO). Our primary contribution involves reformulating the degree selection problem as a QUBO task, reducing the complexity from $O(D^N)$ to a single optimization step pe

H. Abele, P. Ajello, A. Armatol, B. Arnold

The CRAB (Calibrated nuclear Recoils for Accurate Bolometry) project aims to precisely characterize the response of cryogenic detectors to sub-keV nuclear recoils of direct interest for coherent neutrino-nucleus scattering and dark matter search experiments. The CRAB method relies on the radiative capture of thermal neutrons in the target detector, resulting

Mizuki Ono, Takashi Okamoto, Shin'ichiro Ando, Tomoaki Ishiyama

The Cold Dark Matter (CDM) model successfully explains large-scale structure formation, but challenges remain at smaller scales, leading to interest in Warm Dark Matter (WDM) as an alternative. The abundance of Milky Way subhalos depends on the mass of WDM particles, allowing constraints to be obtained by comparing observations and theoretical models. Howeve

Gregorio Falqui, Eleonora Sforza

We apply Poisson reduction techniques to describe asymptotic fully nonlinear models of 2-layer sharply stratified fluids in the Hamiltonian framework. We start by considering the Benjamin Hamiltonian formalism for a stably stratified 2D Euler fluid in a channel of finite height. We use a Marsden-Ratiu reduction scheme for sharply stratified fluids to obtain

Jianing Li

Let $p$ be a prime and let $F$ be a number field. Consider a Galois extension $K/F$ with Galois group $H\rtimes \Delta$ where $H\cong \mathbb{Z}_p$ or $\mathbb{Z}/p^d\mathbb{Z}$, and $\Delta$ is an arbitrary Galois group. The subfields fixed by $H^{p^n} \rtimes \Delta$ $(n=0,1,\cdots)$ form a tower which we call it a potential cyclic $p$-tower in this paper.

Sungwon Park, Dongjoon Lee, Kyeongjin Ahn, Yubin Choi

International aid is a critical mechanism for promoting economic growth and well-being in developing nations, supporting progress toward the Sustainable Development Goals (SDGs). However, tracking aid contributions remains challenging due to labor-intensive data management, incomplete records, and the heterogeneous nature of aid data. Recognizing the urgency

Yisi Luo, Xile Zhao, Deyu Meng

Recently, continuous representation methods emerge as novel paradigms that characterize the intrinsic structures of real-world data through function representations that map positional coordinates to their corresponding values in the continuous space. As compared with the traditional discrete framework, the continuous framework demonstrates inherent superior

Christoph Jabs

State-of-the-art Boolean satisfiability (SAT) solvers constitute a practical and competitive approach for solving various real-world problems. To encourage their widespread adoption, the relatively high barrier of entry following from the low level syntax of SAT and the expert knowledge required to achieve tight integration with SAT solvers should be further

Kamil Kołodziejski

This article proposes novel estimation methods for the Matrix Autoregressive (MAR) model, specifically adaptations of the Yule-Walker equations and Burg's method, addressing limitations in existing techniques. The MAR model, by maintaining a matrix structure and requiring significantly fewer parameters than vector autoregressive (VAR) models, offers a parsim

Yao Du, Huawei Fan, Xingang Wang

A new machine learning scheme, termed versatile reservoir computing, is proposed for sustaining the dynamics of heterogeneous complex networks. We show that a single, small-scale reservoir computer trained on time series from a subset of elements is able to replicate the dynamics of any element in a large-scale complex network, though the elements are of dif

Itsuki Yazawa, Seitaro Yoneda, Akira Furui

Electromyogram (EMG) signals recorded from the skin surface enable intuitive control of assistive devices such as prosthetic limbs. However, in EMG-based motion recognition, collecting comprehensive training data for all target motions remains challenging, particularly for complex combined motions. This paper proposes a method to efficiently recognize combin

Haotian Qin, Dongliang Chang, Yueying Gao, Bingyao Yu

Although existing CLIP-based methods for detecting AI-generated images have achieved promising results, they are still limited by severe feature redundancy, which hinders their generalization ability. To address this issue, incorporating an information bottleneck network into the task presents a straightforward solution. However, relying solely on image-corr

Andy K. Zhang, Joey Ji, Celeste Menders, Riya Dulepet

AI agents have the potential to significantly alter the cybersecurity landscape. Here, we introduce the first framework to capture offensive and defensive cyber-capabilities in evolving real-world systems. Instantiating this framework with BountyBench, we set up 25 systems with complex, real-world codebases. To capture the vulnerability lifecycle, we define

Sangyeon Yoon, Wonje Jeung, Albert No

Large Reasoning Models (LRMs) embed private or copyrighted information not only in their final answers but also throughout multi-step chain-of-thought (CoT) traces, making reliable unlearning far more demanding than in standard LLMs. We introduce Reasoning-TOFU (R-TOFU), the first benchmark tailored to this setting. R-TOFU augments existing unlearning tasks

Sampanna Yashwant Kahu

Efficient task scheduling is paramount in the Linux kernel, where the Completely Fair Scheduler (CFS) meticulously manages CPU resources to balance high utilization with interactive responsiveness. This research pioneers the use of deep learning techniques to predict the sequence of tasks selected by CFS, aiming to evaluate the feasibility of a more generali

Fangxin Liu, Zongwu Wang, JinHong Xia, Junping Zhao

The rapid advancement of large language models (LLMs) has exacerbated the memory bottleneck due to the widening gap between model parameter scaling and hardware capabilities. While post-training quantization techniques effectively reduce memory overhead, existing methods predominantly rely on static quantization strategies, which struggle to adapt to dynamic

Haomin Bai, Dingzhi Yu, Shuai Li, Haipeng Luo

Group distributionally robust optimization (GDRO) aims to develop models that perform well across $m$ distributions simultaneously. Existing GDRO algorithms can only process a fixed number of samples per iteration, either 1 or $m$, and therefore can not support scenarios where the sample size varies dynamically. To address this limitation, we investigate GDR

Yingbo Luo, Meibao Yao, Xueming Xiao

Training a universal controller for robots with different morphologies is a promising research trend, since it can significantly enhance the robustness and resilience of the robotic system. However, diverse morphologies can yield different dimensions of state space and action space, making it difficult to comply with traditional policy networks. Existing met

Pramit Bhattacharyya, Arnab Bhattacharya

Large language models (LLMs) have achieved remarkable success across various natural language processing tasks. However, most LLM models use traditional tokenizers like BPE and SentencePiece, which fail to capture the finer nuances of a morphologically rich language like Bangla (Bengali). In this work, we introduce BanglaByT5, the first byte-level encoder-de

Jie Ma, Ning Qu, Zhitao Gao, Rui Xing

Knowledge graph-based retrieval-augmented generation seeks to mitigate hallucinations in Large Language Models (LLMs) caused by insufficient or outdated knowledge. However, existing methods often fail to fully exploit the prior knowledge embedded in knowledge graphs (KGs), particularly their structural information and explicit or implicit constraints. The fo

Wonje Jeung, Sangyeon Yoon, Hyesoo Hong, Soeun Kim

Large language models (LLMs) are increasingly deployed in real-world applications, raising concerns about the unauthorized use of copyrighted or sensitive data. Machine unlearning aims to remove such 'forget' data while preserving utility and information from the 'retain' set. However, existing evaluations typically assume that forget and retain sets are ful

Wenjie Liu, Zhongliang Liu, Junwei Shu, Changbo Wang

Transferring 2D textures onto complex 3D scenes plays a vital role in enhancing the efficiency and controllability of 3D multimedia content creation. However, existing 3D style transfer methods primarily focus on transferring abstract artistic styles to 3D scenes. These methods often overlook the geometric information of the scene, which makes it challenging

Andronikos Paliathanasis

We investigate hidden symmetries in minimally coupled scalar field cosmology within the FLRW universe, and a perfect fluid with and without interaction to the scalar field. We show that for an exponential potential there exists a set of canonical transformations through which the cosmological field equations can be recast in the form of a free particle in fl

Chi Kit Ng, Long Bai, Guankun Wang, Yupeng Wang

In endoscopic procedures, autonomous tracking of abnormal regions and following circumferential cutting markers can significantly reduce the cognitive burden on endoscopists. However, conventional model-based pipelines are fragile for each component (e.g., detection, motion planning) requires manual tuning and struggles to incorporate high-level endoscopic i

Hyogun Lee, Haksub Kim, Ig-Jae Kim, Yonghun Choi

Video Anomaly Detection (VAD) automatically identifies anomalous events from video, mitigating the need for human operators in large-scale surveillance deployments. However, two fundamental obstacles hinder real-world adoption: domain dependency and real-time constraints -- requiring near-instantaneous processing of incoming video. To this end, we propose Fl

Cristian Di Pietrantonio, Marcin Sokolowski, Christopher Harris, Daniel Price

Petabytes of archival high time resolution observations have been captured with the Murchison Widefield Array. The search for Fast Radio Bursts within these using established software has been limited by its inability to scale on supercomputing infrastructure, necessary to meet the associated computational and memory requirements. Hence, past searches used a

Rina Tazaki, Tomoyuki Akiyama, Akira Furui

Automated epileptic seizure detection from electroencephalogram (EEG) remains challenging due to significant individual differences in EEG patterns across patients. While existing studies achieve high accuracy with patient-specific approaches, they face difficulties in generalizing to new patients. To address this, we propose a detection framework combining

Jiangsheng Huangfu, Zhengyu Song, Tianwei Hou, Anna Li

Fluid antenna systems (FAS) are among the most promising technologies for the sixth generation (6G) mobile communication networks. Unlike traditional fixed-position multiple-input multiple-output (MIMO) systems, a FAS possesses position reconfigurability to switch on-demand among $N$ predefined ports over a prescribed space. This paper explores the performan

Tomas Brauner, Yang Li, Diederik Roest, Tianzhi Wang

We propose a novel type of duality that connects a sequence of well-known theories with even-multiplicity scalar amplitudes: it relates the Yang-Mills theory coupled to a specific scalar matter sector to the nonlinear sigma model on a symmetric coset space, the (multiflavor) Dirac-Born-Infeld theory, and the special Galileon theory. The duality is manifested

Rachana Soni, Navneet Pratap Singh, Neelam Choudhary

In this research article, we design a quantum hash function model from hybrid quantum walks on finite path graph. The hybrid evolution operator consisting of integrated framework of continuous time quantum walks and lackadaisical quantum walks as per choice of bits in binary input message, acts on initial quantum state and generate quantum hash values from p

Pinxin Liu, Haiyang Liu, Luchuan Song, Jason J. Corso

When humans speak, gestures help convey communicative intentions, such as adding emphasis or describing concepts. However, current co-speech gesture generation methods rely solely on superficial linguistic cues (e.g. speech audio or text transcripts), neglecting to understand and leverage the communicative intention that underpins human gestures. This result

Haoyan Yang, Runxue Bao, Cao Xiao, Jun Ma

LLM-as-a-Judge has emerged as a promising tool for automatically evaluating generated outputs, but its reliability is often undermined by potential biases in judgment. Existing efforts to mitigate these biases face key limitations: in-context learning-based methods fail to address rooted biases due to the evaluator's limited capacity for self-reflection, whe

Weiqi Wang, Limeng Cui, Xin Liu, Sreyashi Nag

Goal-oriented script planning, or the ability to devise coherent sequences of actions toward specific goals, is commonly employed by humans to plan for typical activities. In e-commerce, customers increasingly seek LLM-based assistants to generate scripts and recommend products at each step, thereby facilitating convenient and efficient shopping experiences.

Adel Javanmard, Rudrajit Das, Alessandro Epasto, Vahab Mirrokni

Retraining a model using its own predictions together with the original, potentially noisy labels is a well-known strategy for improving the model performance. While prior works have demonstrated the benefits of specific heuristic retraining schemes, the question of how to optimally combine the model's predictions and the provided labels remains largely open

Hana Satou, F Monkey

Domain adaptation remains a challenge when there is significant manifold discrepancy between source and target domains. Although recent methods leverage manifold-aware adversarial perturbations to perform data augmentation, they often neglect precise manifold alignment and systematic exploration of structured perturbations. To address this, we propose GAMA (

Lea Trescher, David Mogalle, Patrick Otto Ludl, Tobias Seidel

This article examines the effect of individual and combined uncertainties in thermodynamic models on the performance of simulated, steady-state Pareto-optimized Dividing Wall Columns. It is a follow-up of the previous work analogously treating deviations in process variables. Such deviations and uncertainties that may even be unknown during the design proces

Fatemeh Ziaeetabar, Florentin Wörgötter

Foundation models have ushered in a new era for multimodal video understanding by enabling the extraction of rich spatiotemporal and semantic representations. In this work, we introduce a novel graph-based framework that integrates a vision-language foundation, leveraging VideoMAE for dynamic visual encoding and BERT for contextual textual embedding, to addr

Hana Satou, Alan Mitkiy, Emma Collins, Finn Kingston

Transfer learning under domain shift remains a fundamental challenge due to the divergence between source and target data manifolds. In this paper, we propose MAADA (Manifold-Aware Adversarial Data Augmentation), a novel framework that decomposes adversarial perturbations into on-manifold and off-manifold components to simultaneously capture semantic variati

Shanshan Pan, Runze Zhang, Yilin Liu, Minglun Gong

The advances in 3D reconstruction technology, such as photogrammetry and LiDAR scanning, have made it easier to reconstruct accurate and detailed 3D models for urban scenes. Nevertheless, these reconstructed models often contain a large number of geometry primitives, making interactive manipulation and rendering challenging, especially on resource-constraine

Yuchao Yi

We study both the Riemannian and Lorentzian Calder\'on problem when a family of Dirichlet-to-Neumann maps are given for an open set of magnetic/electromagnetic potentials. For the Riemannian version, by allowing small perturbations of the magnetic potential, we use the Runge Approximation Theorem to show that the metric can be uniquely determined. There is n

Zhiqing Fang, Xin Liu

Detecting multiple change points in functional data sequences has been increasingly popular and critical in various scientific fields. In this article, we propose a novel two-stage framework for detecting multiple change points in functional data sequences, named as detection by Group Selection and Partial F-test (GS-PF). The detection problem is firstly tra

Dilip Kumar Ghosh, Biswarup Mukhopadhyaya, Sirshendu Samanta, Ritesh K. Singh

The phenomenology of the flipped two-Higgs-doublet model (2HDM) is relatively less explored so far, as compared to the other, commonly discussed, types. It is found that this scenario, like several others, admits of a light neutral pseudoscalar $A$ in the mass range 20 - 60 GeV, consistently with all current experimental data and theoretical constraints. How

Aparajita Dasgupta, Prerna Gulia, Sanjoy Pusti, Sundaram Thangavelu

We prove an analogue of Beurling's theorem on the H-type groups of certain dimensions after establishing the Gutzmer's formula for the H-type groups. We also obtain some other versions of the theorem using the modified Radon transform.

Yifan Liu, Keyu Fan, Weihao Yu, Chenxin Li

Recent advances in generalizable 3D Gaussian Splatting have demonstrated promising results in real-time high-fidelity rendering without per-scene optimization, yet existing approaches still struggle to handle unfamiliar visual content during inference on novel scenes due to limited generalizability. To address this challenge, we introduce MonoSplat, a novel

Yangting Shi, Yinfei Zhu, Renjie He, Le Hui

Omni-domain infrared small target detection (Omni-IRSTD) poses formidable challenges, as a single model must seamlessly adapt to diverse imaging systems, varying resolutions, and multiple spectral bands simultaneously. Current approaches predominantly rely on visual-only modeling paradigms that not only struggle with complex background interference and inher

Jeonghye Kim, Sojeong Rhee, Minbeom Kim, Dohyung Kim

Recent advances in LLM agents have largely built on reasoning backbones like ReAct, which interleave thought and action in complex environments. However, ReAct often produces ungrounded or incoherent reasoning steps, leading to misalignment between the agent's actual state and goal. Our analysis finds that this stems from ReAct's inability to maintain consis

Xuanchi Zhou, Xiaohui Yao, Wentian Lu, Jinjian Guo

Hydrogen-associated filling-controlled Mottronics within electron-correlated system provides a groundbreaking paradigm to explore exotic physical functionality and phenomena. Dynamically controlling hydrogen-induced phase transitions through external fields offers a promising route for designing protonic devices in multidisciplinary fields, but faces high-sp

Jiawei Gu, Ziyue Qiao, Xiao Luo

Graph Neural Networks (GNNs) have shown remarkable performance across various domains, yet they often struggle with model bias, particularly in the presence of class imbalance. This bias can lead to suboptimal performance and unfair predictions, especially for underrepresented classes. We introduce NeuBM (Neutral Bias Mitigation), a novel approach to mitigat

Chaozheng Wang, Zezhou Yang, Shuzheng Gao, Cuiyun Gao

Code completion, a crucial practice in industrial settings, helps developers improve programming efficiency by automatically suggesting code snippets during development. With the emergence of Large Code Models (LCMs), this field has witnessed significant advancements. Due to the natural differences between open-source and industrial codebases, such as coding

Zhehao Huang, Xinwen Cheng, Jie Zhang, Jinghao Zheng

Recent advancements in deep models have highlighted the need for intelligent systems that combine continual learning (CL) for knowledge acquisition with machine unlearning (MU) for data removal, forming the Continual Learning-Unlearning (CLU) paradigm. While existing work treats CL and MU as separate processes, we reveal their intrinsic connection through a

Jiawei Gu, Ziyue Qiao, Zechao Li

The task of graph-level out-of-distribution (OOD) detection is crucial for deploying graph neural networks in real-world settings. In this paper, we observe a significant difference in the relationship between the largest and second-largest eigenvalues of the Laplacian matrix for in-distribution (ID) and OOD graph samples: \textit{OOD samples often exhibit a

Qian Zhou, Xianda Guo, Jilong Wang, Chuanfu Shen

Generalized gait recognition remains challenging due to significant domain shifts in viewpoints, appearances, and environments. Mixed-dataset training has recently become a practical route to improve cross-domain robustness, but it introduces underexplored issues: 1) inter-dataset supervision conflicts, which distract identity learning, and 2) redundant or n

Donald Flynn, Diego Granziol

Backdoor and data-poisoning attacks can flip predictions with tiny training corruptions, yet a sharp theory linking poisoning strength, overparameterization, and regularization is lacking. We analyze ridge least squares with an unpenalized intercept in the high-dimensional regime \(p,n\to\infty\), \(p/n\to c\). Targeted poisoning is modelled by shifting a \(

Bo-Han Lai, Pin-Han Huang, Bo-Han Kung, Shang-Tse Chen

Lipschitz neural networks are well-known for providing certified robustness in deep learning. In this paper, we present a novel, efficient Block Reflector Orthogonal (BRO) layer that enhances the capability of orthogonal layers on constructing more expressive Lipschitz neural architectures. In addition, by theoretically analyzing the nature of Lipschitz neur

Zhipei Xu, Xuanyu Zhang, Qing Huang, Xing Zhou

Recent advances in Artificial Intelligence Generated Content have led to highly realistic synthetic videos, particularly in human-centric scenarios involving speech, gestures, and full-body motion, posing serious threats to information authenticity and public trust. Unlike DeepFake techniques that focus on localized facial manipulation, human-centric video g

Xinran Wang, Muxi Diao, Yuanzhi Liu, Chunyu Wang

Training text-to-image (T2I) models with detailed captions can significantly improve their generation quality. Existing methods often rely on simplistic metrics like caption length to represent the detailness of the caption in the T2I training set. In this paper, we propose a new metric to estimate caption detailness based on two aspects: image coverage rate

Raveena Prasad, Aarush Roy, Suchi Kumari

Assigning tasks efficiently in cloud computing is a challenging problem and is considered an NP-hard problem. Many researchers have used metaheuristic algorithms to solve it, but these often struggle to handle dynamic workloads and explore all possible options effectively. Therefore, this paper presents a new hybrid method that combines two popular algorithm

Lukasz Marciniak, Maja Szymczak, Przemyslaw Wozny, Marcin Runowski

Luminescence thermometry and manometry are exponentially growing areas dealing with the optical detection of temperature and pressure, respectively, being appealing alternatives for conventional thermometers and manometers. The main benefit of luminescent thermometers and manometers is a possibility of remote temperature and/or pressure monitoring, in contra

Yeheng Tong

In this paper, we consider a symmetric teleparallel gravity model that extends the general relativity equivalent model by several parity violating interactions between the gravitational field and a scalar field. We derive three different families of background solutions in flat FRW universe, with three classes of different connections. Through investigations

Giovanni Micheli, Maria Teresa Vespucci, Gianluigi Migliavacca, Dario Siface

The current massive installation of distributed resources in electricity distribution systems is transforming these systems into active dispatching subjects. At the same time, the need to compensate for the intermittent generation of an increasing amount of renewable sources creates the need to acquire more ancillary services. Flexible resources in the distr

Giovanni Micheli, Laureano F. Escudero, Francesca Maggioni, Guzin Bayraksan

In this paper we address the challenge of designing optimal domestic renewable energy systems under multiple sources of uncertainty appearing at different time scales. Long-term uncertainties, such as investment and maintenance costs of different technologies, are combined with short-term uncertainties, including solar radiation, electricity prices, and unco

Wenzhang Tian, Hao Chen, Mingyuan Zhang, Zengqi Chen

Miniaturized spectrometers employing chip solutions are essential for a wide range of applications, such as wearable health monitoring, biochemical sensing, and portable optical coherence tomography. However, the development of integrated spectrometers is hampered by the inherent trade-off between bandwidth-to-resolution, footprint, sampling channels, and op

Iacopo Carusotto

We review the basic concepts of quantum fluids of light and the different techniques that have been developed to exploit driving and dissipation to stabilize and manipulate interesting many-body states. In the weakly interacting regime, this approach has allowed to study, among other, superfluid light, non-equilibrium Bose-Einstein condensation, photonic ana

Olcay Coşkun, Ruslan Muslumov

We generalize Bouc's construction of orthogonal idempotents in the double Burnside algebra to the setting of the double $\mathbb{C}^\times$-fibered Burnside algebra. This yields a structural decomposition of the evaluations of $\mathbb{C}^\times$-fibered biset functors on finite groups. We then construct a complete set of orthogonal idempotents in the catego

Hannah R. Nolasco, Andrew Vargo, Koichi Kise

Self-tracking technologies and wearables automate the process of data collection and insight generation with the support of artificial intelligence systems, with many emerging studies exploring ways to evolve these features further through large-language models (LLMs). This is done with the intent to reduce capture burden and the cognitive stress of health-b

Bushra Ali

Event-by-event fluctuations of the mean transverse momentum ($p_{\rm T}$) of relativistic charged particles are analyzed using the two-particle correlator $\sqrt{C_m}/M(p_{\rm T})_m$, which quantifies the correlations strength in units of the mean $p_{\rm T}$ in proton-proton collision at $\sqrt{s} = 13$ TeV in ALICE both for minimum bias and and high-multip

Jaeyeon Lee, Dong-Wan Choi

Although Vision Transformers (ViTs) have become the standard architecture in computer vision, their massive sizes lead to significant computational overhead. Token compression techniques have attracted considerable attention to address this issue, but they often suffer from severe information loss, requiring extensive additional training to achieve practical

Alice Garbagnati, Cecília Salgado

We consider the countably many families $\mathcal{L}_d$, $d\in\mathbb{N}_{\geq 2}$, of K3 surfaces admitting an elliptic fibration with positive Mordell--Weil rank. We prove that the elliptic fibrations on the very general member of these families have the potential Mordell--Weil rank jump property for $d\neq 2,3$ and moreover the Mordell--Weil rank jump pro

Yunsheng Ma, Burhaneddin Yaman, Xin Ye, Mahmut Yurt

Recent advances have explored integrating large language models (LLMs) into end-to-end autonomous driving systems to enhance generalization and interpretability. However, most existing approaches are limited to either driving performance or vision-language reasoning, making it difficult to achieve both simultaneously. In this paper, we propose ALN-P3, a unif

Mohit Sharma, Adam Fishman, Vikash Kumar, Chris Paxton

Robots in the real world need to perceive and move to goals in complex environments without collisions. Avoiding collisions is especially difficult when relying on sensor perception and when goals are among clutter. Diffusion policies and other generative models have shown strong performance in solving local planning problems, but often struggle at avoiding

Cheng Guo, Jing Jia, Peng Wang, Jing Zhang

Recommendation as a service has improved the quality of our lives and plays a significant role in variant aspects. However, the preference of users may reveal some sensitive information, so that the protection of privacy is required. In this paper, we propose a privacy-preserving, socialized, recommendation protocol that introduces information collected from

Yuante Li, Xu Yang, Xiao Yang, Minrui Xu

Financial markets pose fundamental challenges for asset return prediction due to their high dimensionality, non-stationarity, and persistent volatility. Despite advances in large language models and multi-agent systems, current quantitative research pipelines suffer from limited automation, weak interpretability, and fragmented coordination across key compon

Jinghui Lu, Haiyang Yu, Siliang Xu, Shiwei Ran

Recent advancements in reasoning have significantly enhanced the capabilities of Large Language Models (LLMs) and Multimodal Large Language Models (MLLMs) across diverse tasks. However, excessive reliance on chain-of-thought (CoT) reasoning can impair model performance and brings unnecessarily lengthened outputs, reducing efficiency. Our work reveals that pr

Kunyang Sun, Matthew Du, Joel Yuen-Zhou

The strong coupling between molecules and photonic modes in a Fabry-P\'{e}rot optical cavity, which forms hybrid light-matter states called polaritons, has been demonstrated as a promising route to control the rates of chemical reactions. However, theoretical studies, which largely employ models with a single cavity mode, cannot explain the experimentally ob

Nanxu Gong, Zijun Li, Sixun Dong, Haoyue Bai

Feature Transformation (FT) crafts new features from original ones via mathematical operations to enhance dataset expressiveness for downstream models. However, existing FT methods exhibit critical limitations: discrete search struggles with enormous combinatorial spaces, impeding practical use; and continuous search, being highly sensitive to initialization

Aobo Liang, Yan Sun, Xiaohou Shi, Ke Li

In the past few years, time series foundation models have achieved superior predicting accuracy. However, real-world time series often exhibit significant diversity in their temporal patterns across different time spans and domains, making it challenging for a single model architecture to fit all complex scenarios. In addition, time series data may have mult

Long Wang, Jiongzhi Zheng, Zhengda Xiong, ChuMin Li

Algorithms designed for routing problems typically rely on high-quality candidate edges to guide their search, aiming to reduce the search space and enhance the search efficiency. However, many existing algorithms, like the classical Lin-Kernighan-Helsgaun (LKH) algorithm for the Traveling Salesman Problem (TSP), often use predetermined candidate edges that

Jin Sun, Xinmin Hou

The deficiency of a graph $G$, denoted by $\kd(G)$, is the number of vertices not saturated by a maximum matching. A bone $B_i$ is the tree obtained by attaching two pendent edges to each of the end vertices of a path $P_{i}$. The local independence number of $G$, denoted by $\alpha_l(G)$, is defines as the maximum integer $t$ such that $G$ contains an induc

Litao Ye, Bin Chen, Shrivastava Shivanshu, Chen Sun

Centralized Dynamic Spectrum Sharing (DSS) faces challenges like data security, high management costs, and limited scalability. To address these issues, a blockchain-based DSS scheme has been proposed in this paper. First, we utilize the ERC4907 standard to mint Non-Fungible Spectrum Tokens (NFSTs) that serve as unique identifiers for spectrum resources and

Quanwei Liu, Tao Huang, Jiaqi Yang, Wei Xiang

Semantic segmentation (SS) of RSIs enables the fine-grained interpretation of surface features, making it a critical task in RS analysis. With the increasing diversity and volume of RSIs collected by sensors on various platforms, traditional processing methods struggle to maintain efficiency and accuracy. In response, deep learning (DL) has emerged as a tran

Lanxiang Hu, Mingjia Huo, Yuxuan Zhang, Haoyang Yu

Playing video games requires perception, memory, and planning, exactly the faculties modern large language model (LLM) agents are expected to master. We study the major challenges in using popular video games to evaluate modern LLMs and find that directly dropping LLMs into games cannot make an effective evaluation, for three reasons -- brittle vision percep

Xinran Wang, Songyu Xu, Xiangxuan Shan, Yuxuan Zhang

Cinematography is a cornerstone of film production and appreciation, shaping mood, emotion, and narrative through visual elements such as camera movement, shot composition, and lighting. Despite recent progress in multimodal large language models (MLLMs) and video generation models, the capacity of current models to grasp and reproduce cinematographic techni

Yuta Hamada, Arata Ishige, Yuichi Koga

We determine the maximal gauge groups arising in $E_8$ heterotic string theory on $T^2$. Our analysis proceeds along two approaches. First, we start the moduli space of the supersymmetric heterotic string theory on $T^2$ focusing on points of maximal gauge enhancement. At these special points, the charge lattice can exhibit a $\mathbb{Z}_2$ outer automorphis

Weiqin Chen, Xinjie Zhang, Dharmashankar Subramanian, Santiago Paternain

Transformer models (TMs) have exhibited remarkable in-context reinforcement learning (ICRL) capabilities, allowing them to generalize to and improve in previously unseen environments without re-training or fine-tuning. This is typically accomplished by imitating the complete learning histories of a source RL algorithm over a substantial amount of pretraining

Bowen Liu, Mao Sheng

In this paper we complete the study of the Lan-Sheng-Zuo conjecture proposed in arXiv:1210.8280 for the curve case. Precisely, we prove that every semistable Higgs bundle is strongly semistable for curves of genus $g\leq 1$, and over any curves of genus $g\ge2$ construct explicit examples of semistable Higgs bundles of arbitrary big rank (the first example i

Yunlong Hou, Fengzhuo Zhang, Cunxiao Du, Xuan Zhang

Speculative decoding has emerged as a popular method to accelerate the inference of Large Language Models (LLMs) while retaining their superior text generation performance. Previous methods either adopt a fixed speculative decoding configuration regardless of the prefix tokens, or train draft models in an offline or online manner to align them with the conte

Tong Cheng, Jie Fu, Xinpeng Ling, Huifa Li

Graph Neural Networks (GNNs) have been widely used for graph analysis. Federated Graph Learning (FGL) is an emerging learning framework to collaboratively train graph data from various clients. Although FGL allows client data to remain localized, a malicious server can still steal client private data information through uploaded gradient. In this paper, we f

Badhan Mazumder, Lei Wu, Vince D. Calhoun, Dong Hye Ye

Gaining insights into the structural and functional mechanisms of the brain has been a longstanding focus in neuroscience research, particularly in the context of understanding and treating neuropsychiatric disorders such as Schizophrenia (SZ). Nevertheless, most of the traditional multimodal deep learning approaches fail to fully leverage the complementary

Yang Xu, Swetha Ganesh, Washim Uddin Mondal, Qinbo Bai

This paper investigates infinite-horizon average reward Constrained Markov Decision Processes (CMDPs) with general parametrization. We propose a Primal-Dual Natural Actor-Critic algorithm that adeptly manages constraints while ensuring a high convergence rate. In particular, our algorithm achieves global convergence and constraint violation rates of $\tilde{

Seongmin Hwang, Daeyoung Han, Moongu Jeon

Multispectral object detection aims to leverage complementary information from visible (RGB) and infrared (IR) modalities to enable robust performance under diverse environmental conditions. Our key insight, derived from wavelet analysis and empirical observations, is that IR images contain structurally rich high-frequency information critical for object det

Kunyang Huang, Bin Hu

The rapid advancement of artificial intelligence (AI) has enabled sophisticated audio generation and voice cloning technologies, posing significant security risks for applications reliant on voice authentication. While existing datasets and models primarily focus on distinguishing between human and fully synthetic speech, real-world attacks often involve aud

Badhan Mazumder, Ayush Kanyal, Lei Wu, Vince D. Calhoun

Clinical studies reveal disruptions in brain structural connectivity (SC) and functional connectivity (FC) in neuropsychiatric disorders such as schizophrenia (SZ). Traditional approaches might rely solely on SC due to limited functional data availability, hindering comprehension of cognitive and behavioral impairments in individuals with SZ by neglecting th

Shivam Agarwal, Zimin Zhang, Lifan Yuan, Jiawei Han

Entropy minimization (EM) trains the model to concentrate even more probability mass on its most confident outputs. We show that this simple objective alone, without any labeled data, can substantially improve large language models' (LLMs) performance on challenging math, physics, and coding tasks. We explore three approaches: (1) EM-FT minimizes token-level

Haiduo Huang, Jiangcheng Song, Yadong Zhang, Pengju Ren

Recent advances in knowledge distillation have emphasized the importance of decoupling different knowledge components. While existing methods utilize momentum mechanisms to separate task-oriented and distillation gradients, they overlook the inherent conflict between target-class and non-target-class knowledge flows. Furthermore, low-confidence dark knowledg

Adib Bazgir, Rama chandra Praneeth Madugula, Yuwen Zhang

We introduce a multicrossmodal LLM-agent framework motivated by the growing volume and diversity of materials-science data ranging from high-resolution microscopy and dynamic simulation videos to tabular experiment logs and sprawling literature archives. While recent AI efforts have accelerated individual tasks such as property prediction or image classifica

Yongsheng Song, Zeyu Yang

In this paper, we study the infinite-time mean field games with discounting, establishing an equilibrium where individual optimal strategies collectively regenerate the mean-field distribution. To solve this problem, we partition all agents into a representative player and the social equilibrium. When the optimal strategy of the representative player has the

Zhenyan Yao, Miao Zhang, Lanhu Wu, Yongri Piao

Perturbation with diverse unlabeled data has proven beneficial for semi-supervised medical image segmentation (SSMIS). While many works have successfully used various perturbation techniques, a deeper understanding of learning perturbations is needed. Excessive or inappropriate perturbation can have negative effects, so we aim to address two challenges: how

Sajjad Ghiasvand, Haniyeh Ehsani Oskouie, Mahnoosh Alizadeh, Ramtin Pedarsani

Vision-Language Models (VLMs) such as CLIP have shown remarkable performance in cross-modal tasks through large-scale contrastive pre-training. To adapt these large transformer-based models efficiently for downstream tasks, Parameter-Efficient Fine-Tuning (PEFT) techniques like (Low-Rank Adaptation) LoRA have emerged as scalable alternatives to full fine-tun

Franc Forstneric

We show that biholomorphic maps between certain pairs of Runge domains in the complex affine space $\mathbb C^n$, $n>1$, are limits of holomorphic automorphisms of $\mathbb C^n$. A similar result holds for volume preserving maps and also in Stein manifolds with the density property. This generalizes several results in the literature and provides a considerab