March 2025 arXiv papers — page 189

Xiaotong Huang, He Zhu, Zihan Liu, Weikai Lin

3D Gaussian Splatting (3DGS) has become a crucial rendering technique for many real-time applications. However, the limited hardware resources on today's mobile platforms hinder these applications, as they struggle to achieve real-time performance. In this paper, we propose SeeLe, a general framework designed to accelerate the 3DGS pipeline for resource-cons

Xinhan Zheng, Xueting Wang, Ruotai Li, Huyu Wu

Traditional Chinese medicine (TCM) exhibits remarkable therapeutic efficacy in healthcare through patient-specific formulas. However, current AI-based TCM formula recommendation models and methods mainly focus on data-based textual associations between symptoms and herbs, and have not fully utilized their features and relations at different scales, especiall

Xiutao Zhu, Xiaolin Wang, Yanbo Zhang, Fangfang Zhang

The Tur\'an number $\ex(n,H)$ is the maximum number of edges that an $n$-vertex $H$-free graph can have. The suspension $\widehat{H}$ is obtained from $H$ by adding a new vertex which is adjacent to all vertices of $H$ and a tree is balanced if the sizes of its two color classes differ at most $1$. In this paper, we obtain a sharp bound of $\ex(n,\widehat{T}

Igor V. Bondarev, Alexandra Boltasseva, Jacob B. Khurgin, Vladimir M. Shalaev

Wigner crystallization of free electrons at room temperature is explored for a new class of metallic ultrathin (transdimensional) materials whose properties can be controlled by their thickness. Our calculations of the critical electron density, temperature and the melting curve show that by reducing the material thickness one can Wigner-crystallize free ele

Shanhe You, Xuewen Luo, Xinhe Liang, Jiashu Yu

Evaluation methods for autonomous driving are crucial for algorithm optimization. However, due to the complexity of driving intelligence, there is currently no comprehensive evaluation method for the level of autonomous driving intelligence. In this paper, we propose an evaluation framework for driving behavior intelligence in complex traffic environments, a

Weihua Sun, Zhaonian Zou

Verifying the serializability of transaction histories is essential for users to know if the DBMS ensures the claimed serializable isolation level without potential bugs. Black-box serializability verification is a promising approach. Existing verification methods often have one or more limitations such as incomplete detection of data anomalies, long verific

Chao Zhang, Yifeng Zhou, Shuheng Wang, Wenfa Li

We have recently seen great progress in 3D scene reconstruction through explicit point-based 3D Gaussian Splatting (3DGS), notable for its high quality and fast rendering speed. However, reconstructing dynamic scenes such as complex human performances with long durations remains challenging. Prior efforts fall short of modeling a long-term sequence with dras

Zheng Zhou, Zhe Li, Bo Yu, Lina Hu

The automatic reconstruction of 3D computer-aided design (CAD) models from CAD sketches has recently gained significant attention in the computer vision community. Most existing methods, however, rely on vector CAD sketches and 3D ground truth for supervision, which are often difficult to be obtained in industrial applications and are sensitive to noise inpu

Zhen Wan, Lulu Fan, Xuzhi Li, Xu Kong

We carry out an imaging survey of six globular clusters (GCs) with a limit magnitude to 22 mag at the 5 sigma level, down to the main sequence stars of the respective cluster, as one of the pilot observing program of the Wide Field Survey Telescope (WFST). This paper present the early results of this survey, where we investigate the tidal characters at the p

Cristina Anton, Iain Smith

We propose a method, funWeightClust, based on a family of parsimonious models for clustering heterogeneous functional linear regression data. These models extend cluster weighted models to functional data, and they allow for multivariate functional responses and predictors. The proposed methodology follows the approach used by the the functional high dimensi

Ruisong Xia, Hao Liu, Yongquan Xue

The X-ray quasi-periodic oscillation (QPO) is a remarkable form of variability in systems of compact object accretion. RE J1034+396, harboring the most significant X-ray QPO in active galactic nuclei (AGNs), is the most noteworthy source for in-depth analysis of AGN X-ray QPO properties. A long-term evolution of its QPO has been observed over the course of t

Ruixi Lin, Ziqiao Wang, Yang You

Language models are strong few-shot learners and achieve good overall accuracy in text classification tasks, masking the fact that their results suffer from great class accuracy imbalance. We believe that the pursuit of overall accuracy should not come from enriching the strong classes, but from raising up the weak ones. To address the imbalance, we propose

Junxiang Qiu, Lin Liu, Shuo Wang, Jinda Lu

Feature caching has emerged as an effective strategy to accelerate diffusion transformer (DiT) sampling through temporal feature reuse. It is a challenging problem since (1) Progressive error accumulation from cached blocks significantly degrades generation quality, particularly when over 50\% of blocks are cached; (2) Current error compensation approaches n

Eric B. Kopp, Raymond Kwong

In this paper we develop two axiomatic tests for the controllability of subsystem codes embedded in decoherence-free subspaces of open quantum systems. The tests expand on existing control theory by considering quantum subsystems where a decoherence-protected quantum state is permitted to exit the set of logically encoded states in order to perform a broader

Carlos A. Vital, Román J. Armenta-Rico, Huziel E. Sauceda

Highly accurate force fields are a mandatory requirement to generate predictive simulations. In this regard, Machine Learning Force Fields (MLFFs) have emerged as a revolutionary approach in computational chemistry and materials science, combining the accuracy of quantum mechanical methods with computational efficiency orders of magnitude superior to ab-init

Shuichi Yahagi, Ansei Yonezawa, Hiroki Seto, Heisei Yonezawa

This paper proposes a sparse identification of nonlinear dynamics (SINDy) with control and exogenous inputs for highly accurate and reliable prediction. Although SINDy is recognized as a remarkable approach for identifying nonlinear systems, several challenges remain. Its application to industrial systems remains limited, and multi-step predictions are not g

Yunhao Luo, Utkarsh A. Mishra, Yilun Du, Danfei Xu

Effective trajectory stitching for long-horizon planning is a significant challenge in robotic decision-making. While diffusion models have shown promise in planning, they are limited to solving tasks similar to those seen in their training data. We propose CompDiffuser, a novel generative approach that can solve new tasks by learning to compositionally stit

Kohei Honda, Takeshi Ishita, Yasuhiro Yoshimura, Ryo Yonetani

This paper presents a novel approach to image-goal navigation by integrating 3D Gaussian Splatting (3DGS) with Visual Navigation Models (VNMs), a method we refer to as GSplatVNM. VNMs offer a promising paradigm for image-goal navigation by guiding a robot through a sequence of point-of-view images without requiring metrical localization or environment-specif

Bowen Wu, Wenqing Wang, Haoran Li, Ying Li

Proactive dialogue systems aim to empower chatbots with the capability of leading conversations towards specific targets, thereby enhancing user engagement and service autonomy. Existing systems typically target pre-defined keywords or entities, neglecting user attributes and preferences implicit in dialogue history, hindering the development of long-term us

Rajdeep Roshan Sahu

This research focuses on the development and enhancement of text-to-image denoising diffusion models, addressing key challenges such as limited sample diversity and training instability. By incorporating Classifier-Free Guidance (CFG) and Exponential Moving Average (EMA) techniques, this study significantly improves image quality, diversity, and stability. U

Md-Ferdous Pervej, Patel Pratik, Koushik Manjunatha, Prasad Shamain

Channel models that represent various operating conditions a communication system might experience are important for design and standardization of any communication system. While statistical channel models have long dominated this space, machine learning (ML) is becoming a popular alternative approach. However, existing approaches have mostly focused on pred

Tomoya Naito, Yuto Hijikata, Juzo Zenihiro, Gianluca Colò

We propose a new observable, named the mirror-skin thickness, in order to extract the strength of the charge symmetry breaking (CSB) term in an energy density functional (EDF). The mirror-skin thickness of $ N = 20 $ isotones and $ Z = 20 $ isotopes is studied by using Hartree-Fock-Bogoliubov (HFB) calculations with various Skyrme EDFs and adding CSB and cha

Linqi Ye, Rankun Li, Xiaowen Hu, Jiayi Li

This paper introduces Unity RL Playground, an open-source reinforcement learning framework built on top of Unity ML-Agents. Unity RL Playground automates the process of training mobile robots to perform various locomotion tasks such as walking, running, and jumping in simulation, with the potential for seamless transfer to real hardware. Key features include

Yuhan Yao, Yoshihiko Hasegawa

The barren plateau phenomenon, where the gradients of parametrized quantum circuits become vanishingly small, poses a significant challenge in quantum machine learning. While previous studies attempted to explain the barren plateau phenomenon using the Weingarten formula, the reliance on the Weingarten formula leads to inaccurate conclusions. In this study,

Xiao Luo, Yoshinobu Kuramashi

We investigate the phase diagram of the three-dimensional SU(3) spin model with finite chemical potential, which is an effective Polyakov loop model for finite density QCD, using the tensor renormalization group method. We successfully determine the location of the critical endpoints being free from the complex action problem in the standard Monte Carlo appr

Jinguang Wang, Jingyu Wang, Haifeng Sun, Tingting Yang

Quantization has been widely used to compress and accelerate inference of large language models (LLMs). Existing methods focus on exploring the per-token dynamic calibration to ensure both inference acceleration and model accuracy under 4-bit quantization. However, in autoregressive generation inference of long sequences, the overhead of repeated dynamic qua

Yangjun Sun, Zhiliang Liu

Traditional control methods often show limitations in dealing with complex nonlinear systems, especially when it is difficult to accurately obtain the exact system model, and the control accuracy and stability are difficult to guarantee. To solve this problem, the Koopman operator theory provides an effective method to linearise nonlinear systems, which simp

Wenhao Wang, Zijie Yu, Rui Ye, Jianqing Zhang

Mobile agents have attracted tremendous research participation recently. Traditional approaches to mobile agent training rely on centralized data collection, leading to high cost and limited scalability. Distributed training utilizing federated learning offers an alternative by harnessing real-world user data, providing scalability and reducing costs. Howeve

Tianjun Wei, Wei Wen, Ruizhi Qiao, Xing Sun

Evaluating large language models (LLMs) in diverse and challenging scenarios is essential to align them with human preferences. To mitigate the prohibitive costs associated with human evaluations, utilizing a powerful LLM as a judge has emerged as a favored approach. Nevertheless, this methodology encounters several challenges, including substantial expenses

Yiqing Guo, Nagur Cherukuru, Eric Lehmann, Xiubin Qi

Sea surface temperature (SST) is a fundamental physical parameter characterising the thermal state of sea surface. Due to the intricate thermal interactions between land, sea, and atmosphere, the spatial gradients of SST in coastal waters often appear at finer spatial scales than those in open ocean waters. The Thermal Infrared Sensor (TIRS) onboard Landsat-

Taira Kawamura, Yusuke Kato

We develop a theoretical framework to determine distribution functions in nonequilibrium systems coupled to equilibrium reservoirs, by using the nonequilibrium Green's function technique. As a paradigmatic example, we consider the nonequilibrium distribution function in a nanowire under a bias voltage. We model the system as a tight-binding chain connected t

Junfeng Li, Zengjian Lou, Haixia Yu

In this paper, we investigate the mixed norm estimates for the operator $ T $associated with a dilated plane curve $(ut, u\gamma(t))$, defined by \[ Tf(x, u) := \int_{0}^{1} f(x_1 - ut, x_2 - u\gamma(t)) \, dt, \] where $ x := (x_1, x_2) $ and $\gamma $ is a general plane curve satisfying appropriate smoothness and curvature conditions. More precisely, we es

Ling Team, Binwei Zeng, Chao Huang, Chao Zhang

In this technical report, we tackle the challenges of training large-scale Mixture of Experts (MoE) models, focusing on overcoming cost inefficiency and resource limitations prevalent in such systems. To address these issues, we present two differently sized MoE large language models (LLMs), namely Ling-Lite and Ling-Plus (referred to as "Bailing" in Chinese

Weimin Han, Fang Feng, Fei Wang, Jianguo Huang

Variational-hemivariational inequalities are an important mathematical framework for nonsmooth problems. The framework can be used to study application problems from physical sciences and engineering that involve non-smooth and even set-valued relations, monotone or non-monotone, among physical quantities. Since no analytic solution formulas are expected for

Jingyi Zhao, Jiayu Yang, Haoxiang Yang

Rising labor costs and increasing logistical demands pose significant challenges to modern delivery systems. Automated Electric Vehicles (AEVs) could reduce reliance on delivery personnel and increase route flexibility, but their adoption is limited due to varying customer acceptance and integration complexities. Shared Distribution Locations (SDLs) offer an

Vivek S Borkar, S Sowmya, Raghavendra Tripathi

We recall the classical formulation of PageRank as the stationary distribution of a singularly perturbed irreducible Markov chain that is not irreducible when the perturbation parameter goes to zero. Specifically, we use the Markov chain tree theorem to derive explicit expressions for the PageRank. This analysis leads to some surprising results. These result

Suliman Khan, Sakander Hayat, Mohammed J. F. Alenazi

Let $G^\sigma=(G,\sigma)$ be a connected signed graph and $A(G^\sigma)$ be its adjacency matrix. The positive inertia index of $G^\sigma$, denoted by $p^{+}(G^\sigma)$, is defined as the number of positive eigenvalues of $A(G^\sigma)$. Assume that $G^\sigma$ contains at least one cycle, and let $g_{r}$ be its girth. In this paper, we prove $p^{+}(G^\sigma) \

Anna Lappala

We present a novel phenomenological theory describing how topological constraints in prime-knot ring polymers induce collective (cooperative) modes of motion. In low-complexity knots, chain segments can move quasi-independently. However, as the crossing number increases, the ring's degrees of freedom become collectively coupled: distinct arc segments must mo

Quan Yu, Fangbao Wang, Xin Yuan, Ying Liu

GaAs-based light-emitting diodes (LEDs) are commonly employed in a variety of applications, including medical imaging, biosensing, optical communications, and night vision. In this paper, we present an alternative application of GaAs-based LED with SI-GaAs substrate for X-ray detection and imaging. The mechanism relies on the semiconductor frequency down-con

Hengguang Zhou, Xirui Li, Ruochen Wang, Minhao Cheng

Recently DeepSeek R1 demonstrated how reinforcement learning with simple rule-based incentives can enable autonomous development of complex reasoning in large language models, characterized by the "aha moment", in which the model manifest self-reflection and increased response length during training. However, attempts to extend this success to multimodal rea

Jia-Liang Lu, Mao Song, Peng Wang, Jian-You Guo

We perform a systematic study of the possible molecular states composed of a pair of heavy mesons such as $D^{(*)}D^{(*)}$, $D^{(*)}\bar{D}^{(*)}$ in the framework of the one-boson-exchange model. The exchanged bosons include the pseudoscalar, scalar and vector mesons($\pi$, $\sigma$, $\rho$, $\omega$). We use the Bonn approximation to get the interaction po

Cunchi Lv, Xiao Shi, Zhengyu Lei, Jinyue Huang

Serverless computing, with its ease of management, auto-scaling, and cost-effectiveness, is widely adopted by deep learning (DL) applications. DL workloads, especially with large language models, require substantial GPU resources to ensure QoS. However, it is prone to produce GPU fragments (e.g., 15\%-94\%) in serverless DL systems due to the dynamicity of w

Noah Graham

We calculate the effects of quantum fluctuations of a scalar field in the "ballpoint pen" cosmic string geometry. Using the approach to renormalization established previously for the energy density in two space dimensions, we extend those calculations to $3+1$ dimensions, nonzero scalar mass, and the full stress-energy tensor, including its contribution to t

Jose Ignacio Arroyo, Pablo A. Marquet, Christopher P. Kempes, Geoffrey West

We developed a theory showing that under appropriate normalizations and rescalings, temperature response curves show a remarkably regular behavior and follow a general, universal law. The impressive universality of temperature response curves remained hidden due to various curve-fitting models not well-grounded in first principles. In addition, this framewor

Zeren Chen, Yuenan Hou, Yulin Chen, Li Liu

In this paper, we introduce the HexPlane representation for 3D semantic scene understanding. Specifically, we first design the View Projection Module (VPM) to project the 3D point cloud into six planes to maximally retain the original spatial information. Features of six planes are extracted by the 2D encoder and sent to the HexPlane Association Module (HAM)

Reginald McLean, Evangelos Chatzaroulas, Jordan Terry, Isaac Woungang

Multi-task reinforcement learning (MTRL) aims to endow a single agent with the ability to perform well on multiple tasks. Recent works have focused on developing novel sophisticated architectures to improve performance, often resulting in larger models; it is unclear, however, whether the performance gains are a consequence of the architecture design itself

Apoorva Lal

The use of the two-way fixed effects regression in empirical social science was historically motivated by folk wisdom that it uncovers the Average Treatment effect on the Treated (ATT) as in the canonical two-period two-group case. This belief has come under scrutiny recently due to recent results in applied econometrics showing that it fails to uncover mean

Fshatsion B. Gessesew, Manjith Bose, Kumaravelu Ganesan, Brett C. Johnson

Vanadium silicide, V$_3$Si, is a promising superconductor for silicon-based superconducting (SC) devices due to its compatibility with silicon substrates and its potential for integration into existing semiconductor technologies. However, to date there have been only a limited number of studies of the formation of SC V$_3$Si thin films and the associated str

Gregory R. Werner, Luke C. Adams, John R. Cary

Smoothing short-wavelength charge density variations can stabilize explicit electrostatic particle-in-cell (PIC) plasma simulations against grid heating and cold beam instabilities, which cause unphysical heating when the Debye length is poorly resolved. We demonstrate this by solving the dispersion and by running 1D electrostatic PIC simulations, using an e

Xi Li, Tong Rao, Cihui Pan

Recent feature matching methods have achieved remarkable performance but lack efficiency consideration. In this paper, we revisit the mainstream detector-free matching pipeline and improve all its stages considering both accuracy and efficiency. We propose an Efficient Deep feature Matching network, EDM. We first adopt a deeper CNN with fewer dimensions to e

Alan Frieze, Xavier Perez-Gimenez

We show that w.h.p.\ the random $r$-uniform hypergraph $H_{n,m}$ contains a loose Hamilton cycle, provided $r\geq 3$ and $m\geq \frac{(1+\epsilon)n\log n}{r}$, where $\epsilon$ is an arbitrary positive constant. This is asymptotically best possible, as if $m\leq \frac{(1-\epsilon)n\log n}{r}$ then w.h.p.\ $H_{n,m}$ contains isolated vertices.

Xinghong Mai, Zhao Wang, Lijun Pan, Johannes Schorghuber

We introduce a machine learning molecular dynamics (MLMD) approach to calculate the anharmonic infrared (IR) absorption spectra of polycyclic aromatic hydrocarbons (PAHs), key carriers of interstellar aromatic IR bands. This method accounts for temperature effects in a molecule-specific way and achieves accuracy comparable to conventional quantum chemical ca

Weihao Gao, Zhuo Deng, Zheng Gong, Ziyi Jiang

Insulin resistance (IR) is a key precursor to diabetes and a significant risk factor for cardiovascular disease. Traditional IR assessment methods require multiple blood tests. We developed a simple AI model using only fasting blood glucose to predict IR in non-diabetic populations. Data from the NHANES (1999-2020) and CHARLS (2015) studies were used for mod

Jun-Jie Huang, Zihan Chen, Tianrui Liu, Wentao Zhao

Existing image steganography methods face fundamental limitations in hiding capacity (typically $1\sim7$ images) due to severe information interference and uncoordinated capacity-distortion trade-off. We propose SMILENet, a novel synergistic framework that achieves 25 image hiding through three key innovations: (i) A synergistic network architecture coordina

Shufang Zhang, Jiazheng Wu, Jiacheng He, Kaiyi Wang

This paper presents HyperGraph ROS, an open-source robot operating system that unifies intra-process, inter-process, and cross-device computation into a computational hypergraph for efficient message passing and parallel execution. In order to optimize communication, HyperGraph ROS dynamically selects the optimal communication mechanism while maintaining a c

Changmin Shin, Jaeyong Song, Hongsun Jang, Dogeun Kim

Graph processing requires irregular, fine-grained random access patterns incompatible with contemporary off-chip memory architecture, leading to inefficient data access. This inefficiency makes graph processing an extremely memory-bound application. Because of this, existing graph processing accelerators typically employ a graph tiling-based or processing-in

Elina Y. Zhang, Theron W. Carmichael, Daniel Huber, Keivan G. Stassun

As the intermediate-mass siblings of stars and planets, brown dwarfs (BDs) are vital to study for a better understanding of how objects change across the planet-to-star mass range. Here, we report two low-mass transiting BD systems discovered by TESS, TOI-4776 (TIC 196286578) and TOI-5422 (TIC 80611440), located in an under-populated region of the BD mass-pe

Tong Mu, Yihao Liu, Mehran Armand

Imitation learning frameworks for robotic manipulation have drawn attention in the recent development of language model grounded robotics. However, the success of the frameworks largely depends on the coverage of the demonstration cases: When the demonstration set does not include examples of how to act in all possible situations, the action may fail and can

Jai Geddes Nelson, Xiaochen Liu, Ken Tye Yong

The process of setting up and successfully running Molecular Dynamics Simulations (MDS) is outlined to be incredibly labour and computationally expensive with a very high barrier to entry for newcomers wishing to utilise the benefits and insights of MDS. Here, presented, is a unique Free and Open-Source Software (FOSS) solution that aims to not only reduce t

Chaoran Xiong, Litao Wei, Kehui Ma, Zhen Sun

Event-based visual odometry has recently gained attention for its high accuracy and real-time performance in fast-motion systems. Unlike traditional synchronous estimators that rely on constant-frequency (zero-order) triggers, event-based visual odometry can actively accumulate information to generate temporally high-order estimation triggers. However, exist

Stuart Goldie, Steffan Ott, Anthony Dawson, Tamara Starke

Size selection of liquid-dispersed 2D nanomaterials is a prerequisite for size-dependent studies in earlier stage research and for their targeted application in commercial settings. Centrifugation is the most widespread method for reliably sorting suspensions of polydisperse 2D nanosheets according to size. However, whilst centrifugation is effective, no a p

Weiguang Chen, Junjie Zhang, Jielong Yang, Eng Siong Chng

Array-geometry-agnostic speech separation (AGA-SS) aims to develop an effective separation method regardless of the microphone array geometry. Conventional methods rely on permutation-free operations, such as summation or attention mechanisms, to capture spatial information. However, these approaches often incur high computational costs or disrupt the effect

Jiaying "Lizzy" Liu, Yiheng Su, Praneel Seth

Large language models (LLMs) are increasingly used to assist computational social science research. While prior efforts have focused on text, the potential of leveraging multimodal LLMs (MLLMs) for online video studies remains underexplored. We conduct one of the first case studies on MLLM-assisted video content analysis, comparing AI's interpretations to hu

Shibo Feng, Wanjin Feng, Xingyu Gao, Peilin Zhao

Spiking Neural Networks (SNNs) offer a promising, biologically inspired approach for processing spatiotemporal data, particularly for time series forecasting. However, conventional neuron models like the Leaky Integrate-and-Fire (LIF) struggle to capture long-term dependencies and effectively process multi-scale temporal dynamics. To overcome these limitatio

Wenhao Liang, Wei Zhang, Lin Yue, Miao Xu

Medical image segmentation is fundamental for computer-aided diagnostics, providing accurate delineation of anatomical structures and pathological regions. While common metrics such as Accuracy, DSC, IoU, and HD primarily quantify spatial agreement between predictions and ground-truth labels, they do not assess the calibration quality of segmentation models,

Ruinan Wang, Ian Nabney, Mohammad Golbabaee

Hyperparameter optimization (HPO) is a critical component of machine learning pipelines, significantly affecting model robustness, stability, and generalization. However, HPO is often a time-consuming and computationally intensive task. Traditional HPO methods, such as grid search and random search, often suffer from inefficiency. Bayesian optimization, whil

Muhammad Waqas Shabbir, Sagor Biswas, Rohit Kajla, Sahithi Nadella

Scattering and absorption are fundamental processes in optical engineering and applications. This study investigates the use of the food dye tartrazine to design refractive index fluids that enhance light propagation through scattering media. The optical properties of the solutions were carefully examined using spectrometry and ellipsometry under two extreme

Yifei Gao, Yating Wang, Wing Tat Leung, Zhengya Yang

In this paper, we present a robust and fully discretized method for solving the time fractional diffusion equation with high-contrast multiscale coefficients. We establish the homogenized equation using a multicontinuum approach and employ the exponential integrator method for time discretization. The multicontinuum upscaled model captures the physical chara

Zi-Xuan Zhang, Junsong Cang, Yu Gao, Hong Li

Primordial black holes (PBH) accretion in the late Universe can lead to significant mass growth. A larger mass further accelerates the accretion radiation output for PBHs with initial masses greater than one solar mass, potentially leading to a stringent energy-dumping constraint derived from observations of the cosmic microwave background. The energy inject

Hengrui Xing, Cong Tian, Liang Zhao, Zhi Ma

In recent years, the application of behavioral testing in Natural Language Processing (NLP) model evaluation has experienced a remarkable and substantial growth. However, the existing methods continue to be restricted by the requirements for manual labor and the limited scope of capability assessment. To address these limitations, we introduce AutoTestForge,

Zachary W. Riedel, Mykhaylo Ozerov, Stuart Calder, Daniel P. Shoemaker

Insulators containing chains of magnetic transition metal cations provide platforms for probing spin-$\frac{1}{2}$ dynamics and quantum critical behavior. Li$_2$CoCl$_4$ contains edge-sharing CoCl$_6$ octahedra that form chains along the crystallographic $c$ axis and orders antiferromagnetically at zero field, but questions remain about its applied-field mag

Fathima Shifa M., Shantanu Desai

We implement a search for spatial coincidence between high energy neutrinos detected by the IceCube neutrino detector (using the publicly available 10-year muon track data) and 37 magnetars, including six extragalactic sources. We use the unbinned maximum likelihood method for our analysis. We do not find any such spatial association between any of the known

Renaud Gauthier

We propose the homotopy shape of the Segal topos of derived stacks over simplicial k-algebras as the higher homotopical generalization of the concept of wave function in Quantum Mechanics

Piotr M. Suder, Eric Laber

Information-directed sampling (IDS) is a powerful framework for solving bandit problems which has shown strong results in both Bayesian and frequentist settings. However, frequentist IDS, like many other bandit algorithms, requires that one have prior knowledge of a (relatively) tight upper bound on the norm of the true parameter vector governing the reward

M. G. Cabrera-Padilla, A. Jiménez-Vargas, Takeshi Miura, Moisés Villegas-Vallecillos

Let $A$ be a complex Banach space with a norm $\|f\|=\|f\|_X+\|d(f)\|_Y$ for $f\in A$, where $d$ is a complex linear map from $A$ onto a Banach space $B$, and $\|\cdot\|_K$ represents the supremum norm on a compact Hausdorff space $K$. In this paper, we characterize surjective isometries on $(A,\|\cdot\|)$, which may be nonlinear. This unifies former results

Kaiyu Huang, Hao Wu, Zhubo Shi, Han Zou

Cloud-based Large Language Model (LLM) services often face challenges in achieving low inference latency and meeting Service Level Objectives (SLOs) under dynamic request patterns. Speculative decoding, which exploits lightweight models for drafting and LLMs for verification, has emerged as a compelling technique to accelerate LLM inference. However, existin

Jingyang Liu, Xingyu Zhou, Huajian Ding, Jiaxin Xu

Quantum key distribution (QKD) serves as a cornerstone of secure quantum communication, providing unconditional security grounded in quantum mechanics. While trusted-node networks have facilitated early QKD deployment, their vulnerability to node compromise underscores the need for untrusted-node architectures. Measurement-device-independent QKD (MDI-QKD) an

Takeshi Sakai, Nobuhito Shiomura, Patricio Sanhueza, Kenji Furuya

We have observed the $^{13}$CH$_3$OH $5_1-4_1$ A$^+$, $^{13}$CH$_3$OH $14_1-13_2$ A$^-$, and CH$_2$DOH $8_{2,6}-8_{1,7}$ $e_0$ lines toward 24 high-mass star-forming regions by using Atacama Large Millimeter/submillimeter Array (ALMA) with an angular resolution of about 0$^{\prime\prime}$.3. This resolution corresponds to a linear scale of 400-1600 au, allow

Messi H. J. Lee, Soyeon Jeon, Jacob M. Montgomery, Calvin K. Lai

Current research on bias in Vision Language Models (VLMs) has important limitations: it is focused exclusively on trait associations while ignoring other forms of stereotyping, it examines specific contexts where biases are expected to appear, and it conceptualizes social categories like race and gender as binary, ignoring the multifaceted nature of these id

Adam Labiosa, Josiah P. Hanna

Teams of people coordinate to perform complex tasks by forming abstract mental models of world and agent dynamics. The use of abstract models contrasts with much recent work in robot learning that uses a high-fidelity simulator and reinforcement learning (RL) to obtain policies for physical robots. Motivated by this difference, we investigate the extent to w

Yun Chen, Nuria González-Prelcic, Takayuki Shimizu, Chinmay Mahabal

Channel tracking in millimeter wave (mmWave) vehicular systems is crucial for maintaining robust vehicle-to-infrastructure (V2I) communication links, which can be leveraged to achieve high accuracy vehicle position and orientation tracking as a byproduct of communication. While prior work tends to simplify the system model by omitting critical system factors

Xianjing Dong

How to devise a second main theorem with best error terms is a central problem in the study of Nevanlinna theory. However, it seems difficult to be done for a general non-positively curved K\"ahler manifold. Based on the work of A. Atsuji in Nevanlinna theory, we present a refined form of the second main theorem of meromorphic mappings on a general complete

Lior Gishboliner, Stefan Glock, Peleg Michaeli, Amedeo Sgueglia

Confirming a conjecture of Erd\H{o}s on the chromatic number of Kneser hypergraphs, Alon, Frankl and Lov\'asz proved that in any $q$-colouring of the edges of the complete $r$-uniform hypergraph, there exists a monochromatic matching of size $\lfloor \frac{n+q-1}{r+q-1}\rfloor$. In this paper, we prove a transference version of this theorem. More precisely,

Changhong Lin, Jiarong Lin, Zhiqiang Sui, XiaoZhi Qu

Multi-sensor fusion is essential for autonomous vehicle localization, as it is capable of integrating data from various sources for enhanced accuracy and reliability. The accuracy of the integrated location and orientation depends on the precision of the uncertainty modeling. Traditional methods of uncertainty modeling typically assume a Gaussian distributio

Pei Yang, Qi Tan, Guihua Wen

To remedy the drawbacks of full-mass or fixed-mass constraints in classical optimal transport, we propose adaptive optimal transport which is distinctive from the classical optimal transport in its ability of adaptive-mass preserving. It aims to answer the mathematical problem of how to transport the probability mass adaptively between probability distributi

Anith Selvakumar, Manasa Bharadwaj

Monocular Indoor Semantic Scene Completion (SSC) aims to reconstruct a 3D semantic occupancy map from a single RGB image of an indoor scene, inferring spatial layout and object categories from 2D image cues. The challenge of this task arises from the depth, scale, and shape ambiguities that emerge when transforming a 2D image into 3D space, particularly with

Feng Jiang, Zhiyu Lin, Yiyang Liu, Liumeng Xue

Recent advances in large language models (LLMs) have fundamentally reshaped speech-to-speech (S2S) systems, enabling increasingly natural spoken interaction. However, existing benchmarks still rely heavily on text-based evaluation and largely ignore paralinguistic cues such as prosody, emotion, and speaker traits, which are central to expressive and human-li

Kwang-Il Choe, Dae-Won Choe, Myong Chol Pak

The low Mach number limit for the compressible viscous diffusion approximation model arising in radiation hydrodynamics is rigorously justified. For the 3-D Cauchy problem, the solutions in an equilibrium diffusion regime are shown to converge to the solutions of an incompressible Navier-Stokes equations locally and globally in time as Mach number goes to ze

Hao Zhou, Khalid Zobaid Adnan, Wyatt Allen Jones, Tianli Feng

GaN/AlN interfaces are essential in advanced high-power and high-frequency electronic devices, where effective thermal management is crucial for optimal performance and reliability. This work investigates the thermal boundary conductance (TBC) of standalone and non-standalone GaN/AlN heterostructures using non-equilibrium molecular dynamics (NEMD) driven by

J. C. Geromel, L. Hsu, E. V. L. Nunes

This paper addresses two minimum reaching time control problems within the context of finite stable systems. The well-known Variable Structure Control (VSC) and Unity Vector Control (UVC) strategies are analyzed, with the primary objective of designing optimal and robust state feedback gains that ensure minimum finite time convergence to the origin. This is

Yingji Zhong, Zhihao Li, Dave Zhenyu Chen, Lanqing Hong

Despite recent successes in novel view synthesis using 3D Gaussian Splatting (3DGS), modeling scenes with sparse inputs remains a challenge. In this work, we address two critical yet overlooked issues in real-world sparse-input modeling: extrapolation and occlusion. To tackle these issues, we propose to use a reconstruction by generation pipeline that levera

J. C. Zamora, T. Aumann, S. Bagchi, S. Bishop

\textbf{Background:} Experiments involving nuclear reactions in a storage ring offer exceptional possibilities for precise measurements in inverse kinematics. These experiments provide excellent angular and energy resolution by particle spectroscopy, in addition to high luminosities. However, the extremely low-pressure environment maintained in the storage r

Afroja Akther, Ayesha Arobee, Abdullah Al Adnan, Omum Auyon

As artificial intelligence (AI) systems become increasingly complex and autonomous, concerns over transparency and accountability have intensified. The "black box" problem in AI decision-making limits stakeholders' ability to understand, trust, and verify outcomes, particularly in high-stakes sectors such as healthcare, finance, and autonomous systems. Block

Honglei Lang, Zhangju Liu

Drinfeld classified Poisson homogeneous spaces of a Poisson Lie group in terms of Dirac structures of the Lie bialgebra. In this paper, we study homogeneous spaces of a 2-group and develop Drinfeld theorem in the Poisson 2-group context.

Teng Xiao, Yige Yuan, Mingxiao Li, Zhengyu Chen

This work studies the alignment of large language models with preference data from an imitation learning perspective. We establish a close theoretical connection between reinforcement learning from human feedback RLHF and imitation learning (IL), revealing that RLHF implicitly performs imitation learning on the preference data distribution. Building on this

Hangli Ge, Xiaojie Yang, Jinyu Chen, Francesco Flammini

This paper introduces a traffic evacuation model for railway disruptions to improve resilience. The research focuses on the problem of failure of several nodes or lines on the railway network topology. We proposed a holistic approach that integrates lines of various operator companies as well as external geographical features of the railway system. The optim

Chengwei Zhao, Kun Hu, Jie Xu, Lijun Zhao

The emerging Internet of Things (IoT) applications, such as driverless cars, have a growing demand for high-precision positioning and navigation. Nowadays, LiDAR inertial odometry becomes increasingly prevalent in robotics and autonomous driving. However, many current SLAM systems lack sufficient adaptability to various scenarios. Challenges include decrease

Nils Graef, Andrew Wasielewski

Slim attention shrinks the context memory size by 2x for transformer models with MHA (multi-head attention), which can speed up inference by up to 2x for large context windows. Slim attention is an exact, mathematically identical implementation of the standard attention mechanism and therefore doesn't compromise model accuracy. In other words, slim attention

Christopher J. Lombardi, Alexander G. Kosovichev, Keitarou Matsumoto

The Global Oscillations at Low Frequencies instrument aboard the Solar and Heliospheric Observatory has provided over two decades of continuous, high-precision data, enabling detailed measurements of the Sun's oscillation frequencies. These oscillations, analyzed through Doppler velocity shifts, offer invaluable insights into the Sun's internal structure and

Yang Zhang, Ling-Fang Lin, Adriana Moreo, Satoshi Okamoto

Nickelates have continued to surprise since their unconventional superconductivity was discovered. Recently, the layered nickelate La$_5$Ni$_3$O$_{11}$ with hybrid single-layer and bilayer stacking showed superconductivity under high pressure. This compound combines features of La$_2$NiO$_4$ and La$_3$Ni$_2$O$_7$, but its pairing mechanism remains to be unde