arXiv subjects

Xin Zhang

Publications and source records attributed to Xin Zhang.

At least 343 records · Page 19Linked to original sources

arXiv2025

Interpretable Tile-Based Classification of Paclitaxel Exposure

Medical image analysis is central to drug discovery and preclinical evaluation, where scalable, objective readouts can accelerate decision-making. We address classification of paclitaxel (Taxol) exposure from phase-contrast microscopy of C6 glioma cells -- a task with subtle dose differences that challenges full-image models. We propose a simple tiling-and-aggregation pipeline that operates on local patches and combines tile outputs into an image label, achieving state-of-the-art accuracy on the benchmark dataset and improving over the published baseline by around 20 percentage points, with trends confirmed by cross-validation. To understand why tiling is effective, we further apply Grad-CAM and Score-CAM and attention analyses, which enhance model interpretability and point toward robustness-oriented directions for future medical image research. Code is released to facilitate reproduction and extension.

cs.CV↗

arXiv2025

3D Bayesian Variational Surface Wave Tomography and Application to the Southwest China

Seismic surface wave tomography uses surface wave information to obtain velocity structures in the subsurface. Due to data noise and nonlinearity of the problem, surface wave tomography often has non-unique solutions. It is therefore required to quantify uncertainty of the results in order to better interpret the resulting images. Bayesian inference is the most widely-used method for this purpose. However, the commonly-used Monte Carlo methods require huge computational cost and remains intractable in high-dimensional problems. Variational inference uses optimization to solve Bayesian inverse problems, and therefore can be more efficient in the case of large datasets and high-dimensional parameter spaces. Variational inference has been applied to 2-D surface wave tomographic problems. In this study, we extend the method to 3-D surface wave tomography by directly inverting for 3-D spatial structures from frequency-dependent travel time measurements. Specifically, we apply three variational methods, mean-field automatic differential variational inference (mean-field ADVI), physically structured variational inference (PSVI) and stochastic Stein varational gradient descent (sSVGD) to surface wave tomographic problems using both synthetic data and real data in the Southwest China. The results show that all methods can provide accurate velocity estimates, while sSVGD produces more reasonable uncertainty estimates than mean-field ADVI and PSVI because of Gaussian assumption used in the later methods. In the real data case, the variational methods provide more detailed velocity structures than those obtained using traditional methods, along with reliable uncertainty estimates. We therefore conclude that variational surface wave tomograph can be applied fruitfully to many realistic problems.

physics.geo-ph↗

arXiv2025

Personalized Interpolation: Achieving Efficient Conversion Estimation with Flexible Optimization Windows

Optimizing conversions is crucial in modern online advertising systems, enabling advertisers to deliver relevant products to users and drive business outcomes. However, accurately predicting conversion events remains challenging due to variable time delays between user interactions (e.g., impressions or clicks) and the actual conversions. These delays vary substantially across advertisers and products, necessitating flexible optimization windows tailored to specific conversion behaviors. To address this, we propose a novel \textit{Personalized Interpolation} method that extends existing models based on fixed conversion windows to support flexible advertiser-specific optimization windows. Our method enables accurate conversion estimation across diverse delay distributions without increasing system complexity. We evaluate the effectiveness of the proposed approach through extensive experiments using a real-world ads conversion model. Our results show that this method achieves both high prediction accuracy and improved efficiency compared to existing solutions. This study demonstrates the potential of our Personalized Interpolation method to improve conversion optimization and support a wider range of advertising strategies in large-scale online advertising systems.

cs.LG↗

arXiv2025

R$^2$ec: Towards Large Recommender Models with Reasoning

Large recommender models have extended LLMs as powerful recommenders via encoding or item generation, and recent breakthroughs in LLM reasoning synchronously motivate the exploration of reasoning in recommendation. In this work, we propose R$^2$ec, a unified large recommender model with intrinsic reasoning capability. R$^2$ec introduces a dual-head architecture that supports both reasoning chain generation and efficient item prediction in a single model, significantly reducing inference latency. To overcome the lack of annotated reasoning data, we design RecPO, a reinforcement learning framework that optimizes reasoning and recommendation jointly with a novel fused reward mechanism. Extensive experiments on three datasets demonstrate that R$^2$ec outperforms traditional, LLM-based, and reasoning-augmented recommender baselines, while further analyses validate its competitive efficiency among conventional LLM-based recommender baselines and strong adaptability to diverse recommendation scenarios. Code and checkpoints available at https://github.com/YRYangang/RRec.

cs.IR↗

arXiv2025

PVMark: Enabling Public Verifiability for LLM Watermarking Schemes

Watermarking schemes for large language models (LLMs) have been proposed to identify the source of the generated text, mitigating the potential threats emerged from model theft. However, current watermarking solutions hardly resolve the trust issue: the non-public watermark detection cannot prove itself faithfully conducting the detection. We observe that it is attributed to the secret key mostly used in the watermark detection -- it cannot be public, or the adversary may launch removal attacks provided the key; nor can it be private, or the watermarking detection is opaque to the public. To resolve the dilemma, we propose PVMark, a plugin based on zero-knowledge proof (ZKP), enabling the watermark detection process to be publicly verifiable by third parties without disclosing any secret key. PVMark hinges upon the proof of `correct execution' of watermark detection on which a set of ZKP constraints are built, including mapping, random number generation, comparison, and summation. We implement multiple variants of PVMark in Python, Rust and Circom, covering combinations of three watermarking schemes, three hash functions, and four ZKP protocols, to show our approach effectively works under a variety of circumstances. By experimental results, PVMark efficiently enables public verifiability on the state-of-the-art LLM watermarking schemes yet without compromising the watermarking performance, promising to be deployed in practice.

cs.CR↗

arXiv2025

Detecting stochastic gravitational wave background from cosmic strings with next-generation detector networks: Component separation based on a multi-source astrophysical foreground noise model

Detecting stochastic gravitational wave background (SGWB) from cosmic strings is crucial for unveiling the evolutionary laws of the early universe and validating non-standard cosmological models. This study presents the first systematic evaluation of the detection capabilities of next-generation ground-based gravitational wave detector networks for cosmic strings. By constructing a hybrid signal model incorporating multi-source astrophysical foreground noise, including compact binary coalescences (CBCs) and compact binary hyperbolic encounters (CBHEs), we propose an innovative parameter estimation methodology based on multi-component signal separation. Numerical simulations using one-year observational data reveal three key findings: (1) The CE4020ET network, comprising the Einstein Telescope (ET-10 km) and the Cosmic Explorer (CE-40 km and CE-20 km), achieves nearly one order of magnitude improvement in constraining the cosmic string tension $Gμ$ compared to individual detectors, reaching a relative uncertainty $ΔGμ/ Gμ< 0.5$ for $Gμ> 3.5 \times 10^{-15}$ under standard cosmological framework; (2) The network demonstrates enhanced parameter resolution in non-standard cosmological scenarios, providing a novel approach to probe pre-Big Bang Nucleosynthesis cosmic evolution; (3) Enhanced detector sensitivity amplifies CBHE foreground interference in parameter estimation, while precise modeling of such signals could further refine $Gμ$ constraints by $1-2$ orders of magnitude. This research not only quantifies the detection potential of third-generation detector networks for cosmic string models but also elucidates the intrinsic connection between foreground modeling precision and cosmological parameter estimation accuracy, offering theoretical foundations for optimizing scientific objectives of next-generation gravitational wave observatories.

astro-ph.CO↗

arXiv2025

Process-Level Trajectory Evaluation for Environment Configuration in Software Engineering Agents

Large language model-based agents show promise for software engineering, but environment configuration remains a bottleneck due to heavy manual effort and scarce large-scale, high-quality datasets. Existing benchmarks assess only end-to-end build/test success, obscuring where and why agents succeed or fail. We introduce the Environment Configuration Diagnosis Benchmark, Enconda-bench, which provides process-level trajectory assessment of fine-grained agent capabilities during environment setup-planning, perception-driven error diagnosis, feedback-driven repair, and action to execute final environment configuration. Our task instances are automatically constructed by injecting realistic README errors and are validated in Docker for scalable, high-quality evaluation. Enconda-bench combines process-level analysis with end-to-end executability to enable capability assessments beyond aggregate success rates. Evaluations across state-of-the-art LLMs and agent frameworks show that while agents can localize errors, they struggle to translate feedback into effective corrections, limiting end-to-end performance. To our knowledge, Enconda-bench is the first framework to provide process-level internal capability assessment for environment configuration, offering actionable insights for improving software engineering agents.

cs.SE↗

arXiv2025

A Dual-Branch CNN for Robust Detection of AI-Generated Facial Forgeries

The rapid advancement of generative AI has enabled the creation of highly realistic forged facial images, posing significant threats to AI security, digital media integrity, and public trust. Face forgery techniques, ranging from face swapping and attribute editing to powerful diffusion-based image synthesis, are increasingly being used for malicious purposes such as misinformation, identity fraud, and defamation. This growing challenge underscores the urgent need for robust and generalizable face forgery detection methods as a critical component of AI security infrastructure. In this work, we propose a novel dual-branch convolutional neural network for face forgery detection that leverages complementary cues from both spatial and frequency domains. The RGB branch captures semantic information, while the frequency branch focuses on high-frequency artifacts that are difficult for generative models to suppress. A channel attention module is introduced to adaptively fuse these heterogeneous features, highlighting the most informative channels for forgery discrimination. To guide the network's learning process, we design a unified loss function, FSC Loss, that combines focal loss, supervised contrastive loss, and a frequency center margin loss to enhance class separability and robustness. We evaluate our model on the DiFF benchmark, which includes forged images generated from four representative methods: text-to-image, image-to-image, face swap, and face edit. Our method achieves strong performance across all categories and outperforms average human accuracy. These results demonstrate the model's effectiveness and its potential contribution to safeguarding AI ecosystems against visual forgery attacks.

cs.CV↗

arXiv2025

Prospects for searching for sterile neutrinos in dynamical dark energy cosmologies using joint observations of gravitational waves and $γ$-ray bursts

In the era of third-generation (3G) gravitational-wave (GW) detectors, GW standard siren observations from binary neutron star mergers provide a powerful tool for probing the expansion history of the universe. Since sterile neutrinos can influence cosmic evolution by modifying the radiation content and suppressing structure formation, GW standard sirens offer promising prospects for constraining sterile neutrino properties within a cosmological framework. Building on this, we investigate the prospects for detecting sterile neutrinos in dynamical dark energy (DE) models using joint observations from 3G GW detectors and a future short gamma-ray burst detector, such as a THESEUS-like telescope. We consider three DE models: the $w$CDM, holographic DE (HDE), and Chevallier-Polarski-Linder (CPL) models. Our results show that the properties of DE can influence the constraints on sterile neutrino parameters. Moreover, the inclusion of GW data significantly improves constraints on both sterile neutrino parameters and other cosmological parameters across all three models, compared to the current limits derived from CMB+BAO+SN (CBS) observations. When GW data are included into the CBS dataset, a preference for $ΔN_{\rm eff} > 0$ emerges at approximately the $1σ$ level in the $w$CDM and CPL models, and reaches about $3σ$ in the HDE model. Moreover, the upper limits on $m_{ν,{\rm sterile}}^{\rm eff}$ are reduced by approximately 13%, 75%, and 3% in the $w$CDM, HDE, and CPL models, respectively.

astro-ph.CO↗

arXiv2025

CSST Slitless Spectra: Target Detection and Classification with YOLO

Addressing the spatial uncertainty and spectral blending challenges in CSST slitless spectroscopy, we present a deep learning-driven, end-to-end framework based on the You Only Look Once (YOLO) models. This approach directly detects, classifies, and analyzes spectral traces from raw 2D images, bypassing traditional, error-accumulating pipelines. YOLOv5 effectively detects both compact zero-order and extended first-order traces even in highly crowded fields. Building on this, YOLO11 integrates source classification (star/galaxy) and discrete astrophysical parameter estimation (e.g., redshift bins), showcasing complete spectral trace analysis without other manual preprocessing. Our framework processes large images rapidly, learning spectral-spatial features holistically to minimize errors. We achieve high trace detection precision (YOLOv5) and demonstrate successful quasar identification and binned redshift estimation (YOLO11). This study establishes machine learning as a paradigm shift in slitless spectroscopy, unifying detection, classification, and preliminary parameter estimation in a scalable system. Future research will concentrate on direct, continuous prediction of astrophysical parameters from raw spectral traces.

astro-ph.IM↗

arXiv2025

Natural Language Processing for Cardiology: A Narrative Review

Cardiovascular diseases are becoming increasingly prevalent in modern society, with a profound impact on global health and well-being. These Cardiovascular disorders are complex and multifactorial, influenced by genetic predispositions, lifestyle choices, and diverse socioeconomic and clinical factors. Information about these interrelated factors is dispersed across multiple types of textual data, including patient narratives, medical records, and scientific literature. Natural language processing (NLP) has emerged as a powerful approach for analysing such unstructured data, enabling healthcare professionals and researchers to gain deeper insights that may transform the diagnosis, treatment, and prevention of cardiac disorders. This review provides a comprehensive overview of NLP research in cardiology from 2014 to 2025. We systematically searched six literature databases for studies describing NLP applications across a range of cardiovascular diseases. After a rigorous screening process, we identified 265 relevant articles. Each study was analysed across multiple dimensions, including NLP paradigms, cardiology-related tasks, disease types, and data sources. Our findings reveal substantial diversity within these dimensions, reflecting the breadth and evolution of NLP research in cardiology. A temporal analysis further highlights methodological trends, showing a progression from rule-based systems to large language models. Finally, we discuss key challenges and future directions, such as developing interpretable LLMs and integrating multimodal data. To the best of our knowledge, this review represents the most comprehensive synthesis of NLP research in cardiology to date.

cs.CL↗

arXiv2025

Probing the sign-changeable interaction between dark energy and dark matter with DESI baryon acoustic oscillations and DES supernovae data

There is a possibility of interaction between dark energy and dark matter, and this interaction may also undergo a sign change during the evolution of the universe. In this paper, we utilize the latest observational data to constrain models of a sign-changeable interaction. The data we employ, in addition to the cosmic microwave background data, also encompass the first-year baryon acoustic oscillation data from DESI and the type Ia supernova data of the full 5-year observation from DES. To achieve high generality, we investigate four interacting dark energy (IDE) models with different forms of the interaction term $Q$: (i) IDE1 with $Q = β(a)Hρ_{\rm de}$; (ii) IDE2 with $Q = β(a)Hρ_{\rm c}$; (iii) IDE3 with $Q = β(a)H_0ρ_{\rm de}$; (iv) IDE4 with $Q = β(a)H_0ρ_{\rm c}$. From the analysis, we observe that $β(z) > 0$ at early times and $β(z) < 0$ at late times, with the coupling $β(z)$ crossing the non-interacting line $β(z) = 0$ during cosmic evolution at the 2$σ$ confidence level for the IDE1, IDE3, and IDE4 models. However, for the IDE2 model, $β(z)$ remains consistently negative and does not cross $β(z) = 0$ at the 2$σ$ confidence level. Our findings indicate that the energy transfer is from dark matter to dark energy when dark matter dominates the universe, and from dark energy to dark matter when dark energy dominates, for the IDE1 and IDE3 models. Furthermore, Bayesian evidence suggests that the IDE1 and IDE3 models are moderately preferred over the $Λ$CDM model. The overall outcomes of this study clearly indicate that, based on current observational data, the sign-changeable IDE models are quite compelling and merit further attention.

astro-ph.CO↗

arXiv2025

EDIT: Enhancing Vision Transformers by Mitigating Attention Sink through an Encoder-Decoder Architecture

In this paper, we propose EDIT (Encoder-Decoder Image Transformer), a novel architecture designed to mitigate the attention sink phenomenon observed in Vision Transformer models. Attention sink occurs when an excessive amount of attention is allocated to the [CLS] token, distorting the model's ability to effectively process image patches. To address this, we introduce a layer-aligned encoder-decoder architecture, where the encoder utilizes self-attention to process image patches, while the decoder uses cross-attention to focus on the [CLS] token. Unlike traditional encoder-decoder framework, where the decoder depends solely on high-level encoder representations, EDIT allows the decoder to extract information starting from low-level features, progressively refining the representation layer by layer. EDIT is naturally interpretable demonstrated through sequential attention maps, illustrating the refined, layer-by-layer focus on key image features. Experiments on ImageNet-1k and ImageNet-21k, along with transfer learning tasks, show that EDIT achieves consistent performance improvements over DeiT3 models. These results highlight the effectiveness of EDIT's design in addressing attention sink and improving visual feature extraction.

cs.CV↗

arXiv2025

Supervised Fine-Tuning or Contrastive Learning? Towards Better Multimodal LLM Reranking

In information retrieval, training reranking models mainly focuses on two types of objectives: metric learning (e.g. contrastive loss to increase the predicted scores on relevant query-document pairs) and classification (binary label prediction of relevance vs. irrelevance). For BERT-style encoders, various studies have shown that contrastive learning (CL) can be more effective than discriminative (classification) learning. However, for large language models (LLMs), classification via supervised fine-tuning (SFT), which predicts ''yes'' (resp. ''no'') token for relevant (resp. irrelevant) pairs, appears more promising as it aligns well with the generative nature of LLMs. This divergence raises a central question: which objective is intrinsically better suited to LLM-based reranking, and what mechanism underlies the difference? In this work, we conduct a comprehensive comparison and analysis between CL and SFT for reranking, taking the universal multimodal retrieval (UMR) as the experimental playground. We first decompose the objectives into two components: weight, which controls the magnitude of those updates, and direction, which guides the model updates, then present a unified framework for understanding their interactions. Through probing experiments, we find that SFT provides a substantially stronger weighting scheme than CL, whereas the preferred scoring direction shows no clear winner. Taken together, these results point to a consistent advantage of SFT over CL for LLM reranking. To further validate our findings, we conduct large-scale training with SFT and present new state-of-the-art rerankers on the MRB benchmark. We also provide ablations on SFT settings and expect our findings to benefit future research and applications in this area.

cs.CL↗

arXiv2025

Characterized behaviors of black hole thermodynamics in the supercritical region

The comprehension of universal thermodynamic behaviors in the supercritical region is crucial for examining the characteristics of black hole systems under high temperature and pressure. This study is devoted to the analysis of characteristic lines and crossover behaviors within the supercritical region. By making use of the free energy, we introduce three key thermodynamic quantities: scaled variance, skewness, and kurtosis. Our results demonstrate that the Widom line, associated with the maximal scaled variance, can effectively differentiate between small and large black hole-like subphases, each displaying distinct thermodynamic behaviors within the supercritical region. Furthermore, by utilizing quasinormal modes, we identify the Frenkel line, offering a dynamic perspective to distinguish between small and large black hole-like subphases. These contribute to a deeper comprehension of black hole subphases in the supercritical region, thus illuminating new facets of black hole thermodynamics.

gr-qc↗

arXiv2025

Prospects for probing dark matter particles and primordial black holes with the Square Kilometre Array using the 21 cm power spectrum at cosmic dawn

Probing the nature of dark matter (DM) remains an outstanding problem in modern cosmology. The 21 cm signal, as a sensitive tracer of neutral hydrogen during cosmic dawn, provides a unique means to investigate DM nature during this critical epoch. Annihilation and decay of DM particles, as well as Hawking radiation of primordial black holes (PBHs), can modify the thermal and ionization histories of the early universe, leaving distinctive imprints on the 21 cm power spectrum. Therefore, the redshifted 21 cm power spectrum serves as a powerful tool to investigate such DM processes. In this work, we systematically assess the potential of the upcoming Square Kilometre Array (SKA) to constrain DM and PBH parameters using the 21 cm power spectrum. Assuming $10,000$ hours of integration time, the SKA is projected to reach sensitivities of $\langleσv\rangle \leq 10^{-28}\,{\rm cm}^{3}\,{\rm s}^{-1}$ and $τ\geq 10^{28}\,{\rm seconds}$, for $10\,{\rm GeV}$ DM particles. It can also probe PBHs with masses of $10^{16}\,\mathrm{g}$ and abundances $f_{\mathrm{PBH}} \leq 10^{-6}$. These results indicate that the SKA could place constraints on DM annihilation, decay, and PBH Hawking radiation that are up to two to three orders of magnitude stronger than current limits. Furthermore, the SKA is expected to exceed existing bounds on sub-GeV DM and to probe Hawking radiation from PBHs with masses above $10^{17}\,{\rm g}$, which are otherwise inaccessible by conventional cosmological probes. Overall, the SKA holds great promise for advancing our understanding of both DM particles and PBHs, potentially offering new insights into the fundamental nature of DM.

astro-ph.CO↗

arXiv2025

Slitless Spectroscopy Source Detection Using YOLO Deep Neural Network

Slitless spectroscopy eliminates the need for slits, allowing light to pass directly through a prism or grism to generate a spectral dispersion image that encompasses all celestial objects within a specified area. This technique enables highly efficient spectral acquisition. However, when processing CSST slitless spectroscopy data, the unique design of its focal plane introduces a challenge: photometric and slitless spectroscopic images do not have a one-to-one correspondence. As a result, it becomes essential to first identify and count the sources in the slitless spectroscopic images before extracting spectra. To address this challenge, we employed the You Only Look Once (YOLO) object detection algorithm to develop a model for detecting targets in slitless spectroscopy images. This model was trained on 1,560 simulated CSST slitless spectroscopic images. These simulations were generated from the CSST Cycle 6 and Cycle 9 main survey data products, representing the Galactic and nearby galaxy regions and the high galactic latitude regions, respectively. On the validation set, the model achieved a precision of 88.6% and recall of 90.4% for spectral lines, and 87.0% and 80.8% for zeroth-order images. In testing, it maintained a detection rate >80% for targets brighter than 21 mag (medium-density regions) and 20 mag (low-density regions) in the Galactic and nearby galaxies regions, and >70% for targets brighter than 18 mag in high galactic latitude regions.

astro-ph.IM↗

arXiv2025

AirScape: An Aerial Generative World Model with Motion Controllability

How to enable agents to predict the outcomes of their own motion intentions in three-dimensional space has been a fundamental problem in embodied intelligence. To explore general spatial imagination capability, we present AirScape, the first world model designed for six-degree-of-freedom aerial agents. AirScape predicts future observation sequences based on current visual inputs and motion intentions. Specifically, we construct a dataset for aerial world model training and testing, which consists of 11k video-intention pairs. This dataset includes first-person-view videos capturing diverse drone actions across a wide range of scenarios, with over 1,000 hours spent annotating the corresponding motion intentions. Then we develop a two-phase schedule to train a foundation model--initially devoid of embodied spatial knowledge--into a world model that is controllable by motion intentions and adheres to physical spatio-temporal constraints. Experimental results demonstrate that AirScape significantly outperforms existing foundation models in 3D spatial imagination capabilities, especially with over a 50% improvement in metrics reflecting motion alignment. The project is available at: https://embodiedcity.github.io/AirScape/.

cs.RO↗