arXiv subjects

Xin Wang

Publications and source records attributed to Xin Wang.

At least 739 records · Page 41Linked to original sources

arXiv2025

Restoring Missing Modes of 21cm Intensity Mapping with Deep Learning: Impact on BAO Reconstruction

In 21cm intensity mapping of the large-scale structure (LSS), regions in Fourier space could be compromised by foreground contamination. In interferometric observations, this contamination, known as the foreground wedge, is exacerbated by the chromatic response of antennas, leading to substantial data loss. Meanwhile, the baryonic acoustic oscillation (BAO) reconstruction, which operates in configuration space to "linearize" the BAO signature, offers improved constraints on the sound horizon scale. However, missing modes within these contaminated regions can negatively impact the BAO reconstruction algorithm. To address this challenge, we employ the deep learning model U-Net to recover the lost modes before applying the BAO reconstruction algorithm. Despite hardware limitations, such as GPU memory, our results demonstrate that the AI-restored 21cm temperature map achieves a high correlation with the original signal, with a correlation ratio of approximately $0.9$ at $k \sim 1 h/Mpc$. Furthermore, subsequent BAO reconstruction indicates that the AI restoration has minimal impact on the performance of the `linearized' BAO signal, proving the effectiveness of the machine learning approach to mitigate the impact of foreground contamination. Interestingly, we demonstrate that the AI model trained on coarser fields can be effectively applied to finer fields, achieving even higher correlation. This success is likely attributable to the scale-invariance properties of non-linear mode coupling in large-scale structure and the hierarchical structure of the U-Net architecture.

astro-ph.CO↗

arXiv2025

Recovering Cosmic Structure with a Simple Physical Constraint

Radio observation of the large-scale structure (LSS) of our Universe faces major challenges from foreground contamination, which is many orders of magnitude stronger than the cosmic signal. While other foreground removal techniques struggle with complex systematics, methods like foreground avoidance emerge as effective alternatives. However, this approach inevitably results in the loss of Fourier modes and a reduction in cosmological constraints. We present a novel method that, by enforcing the non-negativity of the observed field in real space, allows us to recover some of the lost information, particularly phase angles. We demonstrate that the effectiveness of this straightforward yet powerful technique arises from the mode mixing from the non-linear evolution of LSS. Since the non-negativity is ensured by mass conservation, one of the key principles of the cosmic dynamics, we can restore the lost modes without explicitly expressing the exact form of the mode mixing. Unlike previous methods, our approach utilizes information from highly non-linear scales, and has the potential to revolutionize the analysis of radio observational data in cosmology. Crucially, we demonstrate that in long-baseline interferometric observations, such as those from the Square Kilometre Array (SKA), it is still possible to recover the baryonic acoustic oscillation (BAO) signature despite not directly covering the relevant scales. This opens up potential future survey designs for cosmological detection.

astro-ph.CO↗

arXiv2025

Tunable coherent microwave beam splitter and combiner at the single-photon level

A beam splitter is a key component used to direct and combine light paths in various optical and microwave systems. It plays a crucial role in devices like interferometers, such as the Mach-Zehnder and Hong-Ou-Mandel setups, where it splits light into different paths for interference measurements. These measurements are vital for precise phase and coherence testing in both classical and quantum optical experiments. In this work, we present a nonlinear beam splitter and beam combiner utilizing a frequency-tunable superconducting artificial atom in a one-dimensional open waveguide. This beam splitter is highly versatile, with adjustable transparency ranging from unity to zero for signals at the single-photon level. Additionally, the beam combiner can merge two coherent beams, generating interference fringes as the relative phase between them varies.

quant-ph↗

arXiv2025

A Multimodal Knowledge-enhanced Whole-slide Pathology Foundation Model

Remarkable strides in computational pathology have been made in the task-agnostic foundation model that advances the performance of a wide array of downstream clinical tasks. Despite the promising performance, there are still several challenges. First, prior works have resorted to either vision-only or image-caption data, disregarding pathology reports with more clinically authentic information from pathologists and gene expression profiles which respectively offer distinct knowledge for versatile clinical applications. Second, the current progress in pathology FMs predominantly concentrates on the patch level, where the restricted context of patch-level pretraining fails to capture whole-slide patterns. Even recent slide-level FMs still struggle to provide whole-slide context for patch representation. In this study, for the first time, we develop a pathology foundation model incorporating three levels of modalities: pathology slides, pathology reports, and gene expression data, which resulted in 26,169 slide-level modality pairs from 10,275 patients across 32 cancer types, amounting to over 116 million pathological patch images. To leverage these data for CPath, we propose a novel whole-slide pretraining paradigm that injects the multimodal whole-slide context into the patch representation, called Multimodal Self-TAught PRetraining (mSTAR). The proposed paradigm revolutionizes the pretraining workflow for CPath, enabling the pathology FM to acquire the whole-slide context. To the best of our knowledge, this is the first attempt to incorporate three modalities at the whole-slide context for enhancing pathology FMs. To systematically evaluate the capabilities of mSTAR, we built the largest spectrum of oncological benchmark, spanning 7 categories of oncological applications in 15 types of 97 practical oncological tasks.

cs.CV↗

arXiv2025

Probing the He II reionization ERa via Absorbing C IV Historical Yield (HIERACHY) III: The C IV absorber catalog and initial results on cosmic abundance evolution at $z\approx 3-5$

As part of the HIERACHY program, we collect the high-SN and high-spectral resolution optical spectra of 25 quasars at $z\approx4-5$ to constrain the C IV evolution at $z\approx 3-5$. In this paper, we report a catalog of 626 (1263) C IV absorption systems (components) at $z\approx3-5$ with a 50% completeness column density of log$(N_{\rm CIV}/\rm cm^{-2}) \approx 12.3$. The HIERACHY/MIKE C IV sample is one of the best C IV absorber samples optimized to study the IGM during the He II reionization epoch. Using 557 (1090) intervening absorption systems (components), we found the column density distribution function of C IV absorption systems with log$(N_{\rm CIV}/\rm cm^{-2})\gtrsim 12.3$ has a broken power-law shape, with the turn-over column density log$(N_{\rm crit}/\rm cm^{-2}) = 13.35^{+0.20}_{-0.19}$, which is close to or smaller than the detection limit of most literature samples. We also found that both comoving path length number density $dn/dX$ and cosmic abundance $Ω$ for C IV absorption systems with log$(N_{\rm CIV}/\rm cm^{-2})> 13.2$ show an increase (at the 2.2$σ$ and 1.4$σ$ levels, respectively) from redshift $z\approx5$ to 3, while absorption systems with log$(N_{\rm CIV}/\rm cm^{-2})= 12.3-13.2$ exhibit a constant $dn/dX$ and $Ω_{\rm CIV}$.

astro-ph.GA↗

arXiv2025

Complementary Advantages: Exploiting Cross-Field Frequency Correlation for NIR-Assisted Image Denoising

Existing single-image denoising algorithms often struggle to restore details when dealing with complex noisy images. The introduction of near-infrared (NIR) images offers new possibilities for RGB image denoising. However, due to the inconsistency between NIR and RGB images, the existing works still struggle to balance the contributions of two fields in the process of image fusion. In response to this, in this paper, we develop a cross-field Frequency Correlation Exploiting Network (FCENet) for NIR-assisted image denoising. We first propose the frequency correlation prior based on an in-depth statistical frequency analysis of NIR-RGB image pairs. The prior reveals the complementary correlation of NIR and RGB images in the frequency domain. Leveraging frequency correlation prior, we then establish a frequency learning framework composed of Frequency Dynamic Selection Mechanism (FDSM) and Frequency Exhaustive Fusion Mechanism (FEFM). FDSM dynamically selects complementary information from NIR and RGB images in the frequency domain, and FEFM strengthens the control of common and differential features during the fusion process of NIR and RGB features. Extensive experiments on simulated and real data validate that the proposed method outperforms other state-of-the-art methods. The code will be released at https://github.com/yuchenwang815/FCENet.

cs.CV↗

arXiv2025

MuMA: 3D PBR Texturing via Multi-Channel Multi-View Generation and Agentic Post-Processing

Current methods for 3D generation still fall short in physically based rendering (PBR) texturing, primarily due to limited data and challenges in modeling multi-channel materials. In this work, we propose MuMA, a method for 3D PBR texturing through Multi-channel Multi-view generation and Agentic post-processing. Our approach features two key innovations: 1) We opt to model shaded and albedo appearance channels, where the shaded channels enables the integration intrinsic decomposition modules for material properties. 2) Leveraging multimodal large language models, we emulate artists' techniques for material assessment and selection. Experiments demonstrate that MuMA achieves superior results in visual quality and material fidelity compared to existing methods.

cs.CV↗

arXiv2025

Optical Spiking Neurons Enable High-Speed and Energy-Efficient Optical Neural Networks

Optical neural networks (ONNs) perform extensive computations using photons instead of electrons, resulting in passively energy-efficient and low-latency computing. Among various ONNs, the diffractive optical neural networks (DONNs) particularly excel in energy efficiency, bandwidth, and parallelism, therefore attract considerable attention. However, their performance is limited by the inherent constraints of traditional frame-based sensors, which process and produce dense and redundant information at low operating frequency. Inspired by the spiking neurons in human neural system, which utilize a thresholding mechanism to transmit information sparsely and efficiently, we propose integrating a threshold-locking method into neuromorphic vision sensors to generate sparse and binary information, achieving microsecond-level accurate perception similar to human spiking neurons. By introducing novel Binary Dual Adaptive Training (BAT) and Optically Parallel Mixture of Experts (OPMoE) inference methods, the high-speed, spike-based diffractive optical neural network (S2NN) demonstrates an ultra-fast operating speed of 3649 FPS, which is 30 fold faster than that of reported DONNs, delivering a remarkable computational speed of 417.96 TOPS and a system energy efficiency of 12.6 TOPS/W. Our work demonstrates the potential of incorporating neuromorphic architecture to facilitate optical neural network applications in real-world scenarios for both low-level and high-level machine vision tasks.

physics.optics↗

arXiv2025

SVD-LLM: Truncation-aware Singular Value Decomposition for Large Language Model Compression

The advancements in Large Language Models (LLMs) have been hindered by their substantial sizes, which necessitates LLM compression methods for practical deployment. Singular Value Decomposition (SVD) offers a promising solution for LLM compression. However, state-of-the-art SVD-based LLM compression methods have two key limitations: truncating smaller singular values may lead to higher compression loss, and the lack of update on the compressed weights after SVD truncation. In this work, we propose SVD-LLM, a SVD-based post-training LLM compression method that addresses the limitations of existing methods. SVD-LLM incorporates a truncation-aware data whitening technique to ensure a direct mapping between singular values and compression loss. Moreover, SVD-LLM adopts a parameter update with sequential low-rank approximation to compensate for the accuracy degradation after SVD compression. We evaluate SVD-LLM on 10 datasets and seven models from three different LLM families at three different scales. Our results demonstrate the superiority of SVD-LLM over state-of-the-arts, especially at high model compression ratios. Our code is available at https://github.com/AIoT-MLSys-Lab/SVD-LLM

cs.CL↗

arXiv2025

SVD-LLM V2: Optimizing Singular Value Truncation for Large Language Model Compression

Despite significant advancements, the practical deployment of Large Language Models (LLMs) is often hampered by their immense sizes, highlighting the need for effective compression techniques. Singular Value Decomposition (SVD) is a promising LLM compression technique. However, existing SVD-based compression methods fall short in reducing truncation losses, leading to less competitive performance in compressed models. In this work, we introduce SVD-LLM V2, a SVD-based LLM compression method that optimizes singular value truncation in SVD compression with two techniques. First, SVD-LLM V2 proposes to use theoretical truncation loss of weight matrices to assign a unique compression ratio to each weight matrix at different layers to accommodate weight redundancy heterogeneity. Second, SVD-LLM V2 proposes loss-optimized weight truncation to ensure that the truncated singular values result in a lower and more stable truncation loss in practice. We evaluate SVD-LLM V2 on ten datasets and five LLMs at various scales. Our results show SVD-LLM V2 outperforms state-of-the-art SVD-based LLM compression methods. Our code is available at https://github.com/AIoT-MLSys-Lab/SVD-LLM

cs.CL↗

arXiv2025

D2O: Dynamic Discriminative Operations for Efficient Long-Context Inference of Large Language Models

Generative inference in Large Language Models (LLMs) is impeded by the growing memory demands of Key-Value (KV) cache, especially for longer sequences. Traditional KV cache eviction strategies, which discard less critical KV pairs based on attention scores, often degrade generation quality, leading to issues such as context loss or hallucinations. In this work, we introduce Dynamic Discriminative Operations (D2O), a KV cache compression method that optimizes KV cache size dynamically and discriminatively at two levels without fine-tuning, while preserving essential context. At layer level, D2O leverages the varying densities of attention weights between shallow and deep layers to dynamically determine which layers should avoid excessive eviction via a novel dynamic allocation strategy to minimize information loss. At token level, D2O incorporates a compensation mechanism that maintains a similarity threshold to re-discriminate the importance of currently discarded tokens, determining whether they should be recalled and merged with similar tokens. We conduct experiments on various benchmarks and LLM architectures. Our results show that D2O not only achieves significant memory savings and enhances inference throughput by more than 3$\times$ but also maintains high-quality long-text generation.

cs.CL↗

arXiv2025

MEDA: Dynamic KV Cache Allocation for Efficient Multimodal Long-Context Inference

Long-context Multimodal Large Language Models (MLLMs) that incorporate long text-image and text-video modalities, demand substantial resources as their multimodal Key-Value (KV) caches grow with increasing input lengths, challenging inference efficiency. Existing methods for KV cache compression, in both text-only and multimodal LLMs, have neglected attention density variations across layers, thus often adopting uniform or progressive reduction strategies for layer-wise cache allocation. In this work, we propose MEDA, a dynamic layer-wise KV cache allocation method for efficient multimodal long-context inference. As its core, MEDA utilizes cross-modal attention entropy to determine the KV cache size at each MLLMs layer. Given the dynamically allocated KV cache size at each layer, MEDA also employs a KV pair selection scheme to identify which KV pairs to select and a KV pair merging strategy that merges the selected and non-selected ones to preserve information from the entire context. MEDA achieves up to 72% KV cache memory reduction and 2.82 times faster decoding speed, while maintaining or enhancing performance on various multimodal tasks in long-context settings, including multi-images and long-video scenarios. Our code is released at https://github.com/AIoT-MLSys-Lab/MEDA.

cs.CL↗

arXiv2025

Behavior Importance-Aware Graph Neural Architecture Search for Cross-Domain Recommendation

Cross-domain recommendation (CDR) mitigates data sparsity and cold-start issues in recommendation systems. While recent CDR approaches using graph neural networks (GNNs) capture complex user-item interactions, they rely on manually designed architectures that are often suboptimal and labor-intensive. Additionally, extracting valuable behavioral information from source domains to improve target domain recommendations remains challenging. To address these challenges, we propose Behavior importance-aware Graph Neural Architecture Search (BiGNAS), a framework that jointly optimizes GNN architecture and data importance for CDR. BiGNAS introduces two key components: a Cross-Domain Customized Supernetwork and a Graph-Based Behavior Importance Perceptron. The supernetwork, as a one-shot, retrain-free module, automatically searches the optimal GNN architecture for each domain without the need for retraining. The perceptron uses auxiliary learning to dynamically assess the importance of source domain behaviors, thereby improving target domain recommendations. Extensive experiments on benchmark CDR datasets and a large-scale industry advertising dataset demonstrate that BiGNAS consistently outperforms state-of-the-art baselines. To the best of our knowledge, this is the first work to jointly optimize GNN architecture and behavior data importance for cross-domain recommendation.

cs.IR↗

arXiv2025

VizTrust: A Visual Analytics Tool for Capturing User Trust Dynamics in Human-AI Communication

Trust plays a fundamental role in shaping the willingness of users to engage and collaborate with artificial intelligence (AI) systems. Yet, measuring user trust remains challenging due to its complex and dynamic nature. While traditional survey methods provide trust levels for long conversations, they fail to capture its dynamic evolution during ongoing interactions. Here, we present VizTrust, which addresses this challenge by introducing a real-time visual analytics tool that leverages a multi-agent collaboration system to capture and analyze user trust dynamics in human-agent communication. Built on established human-computer trust scales-competence, integrity, benevolence, and predictability-, VizTrust enables stakeholders to observe trust formation as it happens, identify patterns in trust development, and pinpoint specific interaction elements that influence trust. Our tool offers actionable insights into human-agent trust formation and evolution in real time through a dashboard, supporting the design of adaptive conversational agents that responds effectively to user trust signals.

cs.HC↗

arXiv2025

GrInAdapt: Scaling Retinal Vessel Structural Map Segmentation Through Grounding, Integrating and Adapting Multi-device, Multi-site, and Multi-modal Fundus Domains

Retinal vessel segmentation is critical for diagnosing ocular conditions, yet current deep learning methods are limited by modality-specific challenges and significant distribution shifts across imaging devices, resolutions, and anatomical regions. In this paper, we propose GrInAdapt, a novel framework for source-free multi-target domain adaptation that leverages multi-view images to refine segmentation labels and enhance model generalizability for optical coherence tomography angiography (OCTA) of the fundus of the eye. GrInAdapt follows an intuitive three-step approach: (i) grounding images to a common anchor space via registration, (ii) integrating predictions from multiple views to achieve improved label consensus, and (iii) adapting the source model to diverse target domains. Furthermore, GrInAdapt is flexible enough to incorporate auxiliary modalities such as color fundus photography, to provide complementary cues for robust vessel segmentation. Extensive experiments on a multi-device, multi-site, and multi-modal retinal dataset demonstrate that GrInAdapt significantly outperforms existing domain adaptation methods, achieving higher segmentation accuracy and robustness across multiple domains. These results highlight the potential of GrInAdapt to advance automated retinal vessel analysis and support robust clinical decision-making.

eess.IV↗

arXiv2025

Quantum Turbulence Across Dimensions: Crossover from two- to three-dimension

We investigate the dynamic transition of quantum turbulence (QT) in a confined potential field as the system evolves from purely two-dimensional (2D) to quasi-two-dimensional, and ultimately to three-dimensional (3D), by fixing the lateral dimensions of the trapping box while varying its height. In the 2DQT, distinct Onsager vortex cluster formation and inverse energy cascade are observed, while 3DQT exhibits a direct energy cascade consistent with the Vinen turbulence decay rate, which display striking differences. By systematically altering the system height, we explore how dimensionality drives the differentiation of turbulence types and find that this transition is closely related to the excitation of Kelvin waves. Kelvin waves not only introduce additional dissipation mechanisms but also serve as mediators for direct energy transfer across scales. When the wavelength of the permitted Kelvin waves exceeds the critical size of vortex clusters, turbulence begins transitioning to 3D type, culminating in fully developed 3DQT at the characteristic scale. In the transitional region, we observe continuous variations in the decay rate and vortex cluster correlation functions.

cond-mat.quant-gas↗

arXiv2025

SED2AM: Solving Multi-Trip Time-Dependent Vehicle Routing Problem using Deep Reinforcement Learning

Deep reinforcement learning (DRL)-based frameworks, featuring Transformer-style policy networks, have demonstrated their efficacy across various vehicle routing problem (VRP) variants. However, the application of these methods to the multi-trip time-dependent vehicle routing problem (MTTDVRP) with maximum working hours constraints -- a pivotal element of urban logistics -- remains largely unexplored. This paper introduces a DRL-based method called the Simultaneous Encoder and Dual Decoder Attention Model (SED2AM), tailored for the MTTDVRP with maximum working hours constraints. The proposed method introduces a temporal locality inductive bias to the encoding module of the policy networks, enabling it to effectively account for the time-dependency in travel distance or time. The decoding module of SED2AM includes a vehicle selection decoder that selects a vehicle from the fleet, effectively associating trips with vehicles for functional multi-trip routing. Additionally, this decoding module is equipped with a trip construction decoder leveraged for constructing trips for the vehicles. This policy model is equipped with two classes of state representations, fleet state and routing state, providing the information needed for effective route construction in the presence of maximum working hours constraints. Experimental results using real-world datasets from two major Canadian cities not only show that SED2AM outperforms the current state-of-the-art DRL-based and metaheuristic-based baselines but also demonstrate its generalizability to solve larger-scale problems.

cs.AI↗

arXiv2025

A Survey on Evaluating Large Language Models in Code Generation Tasks

This paper provides a comprehensive review of the current methods and metrics used to evaluate the performance of Large Language Models (LLMs) in code generation tasks. With the rapid growth in demand for automated software development, LLMs have demonstrated significant potential in the field of code generation. The paper begins by reviewing the historical development of LLMs and their applications in code generation. Next, it details various methods and metrics for assessing the code generation capabilities of LLMs, including code correctness, efficiency, readability, and evaluation methods based on expert review and user experience. The paper also evaluates the widely used benchmark datasets, identifying their limitations and proposing directions for future improvements. Specifically, the paper analyzes the performance of code generation models across different tasks by combining multiple evaluation metrics, such as code compilation/interpretation success rates, unit test pass rates, and performance and efficiency metrics, to comprehensively assess the practical application of LLMs in code generation. Finally, the paper discusses the challenges faced in evaluating LLMs in code generation, particularly how to ensure the comprehensiveness and accuracy of evaluation methods and how to adapt to the evolving practices of software development. These analyses and discussions provide valuable insights for further optimizing and improving the application of LLMs in code generation tasks.

cs.SE↗