arXiv subjects

Zhili Wang

Publications and source records attributed to Zhili Wang.

17 recordsLinked to original sources

arXiv2026

Heavy Seed Black Hole Growth in Metal-Enriched Halos through Disk-Induced Stellar Disruptions: A Semi-Analytical Modelling

Recent simulations suggest that heavy seed black holes may form in weakly metal-enriched atomic cooling halos, where the supermassive-star progenitor and small-scale stellar fragments emerge nearly coevally. In this picture, the newly born heavy seed is naturally embedded in a metal-enriched Pop~I/II nuclear star cluster rather than in an isolated pristine environment. We investigate whether disk-induced tidal disruption events (TDEs) from these Pop~I/II stars can provide an efficient and sustained growth channel for heavy seed black holes. We construct a semi-analytical model for stellar orbital damping, disk capture, migration, and tidal disruption around a heavy seed black hole, and incorporate the resulting disk-induced TDE contribution into cosmological merger trees with baryonic and metallicity evolution. Heavy seed host halos are selected from atomic cooling halos with $Z\lesssim10^{-3}Z_\odot$ that satisfy either a rapid gas-inflow criterion or a strong Lyman--Werner radiation criterion. We find that disk-induced Pop~I/II TDEs can dominate the early growth of heavy seeds: the median black hole mass grows from $\sim10^4\,M_\odot$ to $\sim10^5\,M_\odot$ within the first $\sim0.1$ Gyr after seed formation, and reaches several $10^5\,M_\odot$ by $\sim0.2$ Gyr. The cumulative mass supplied by TDEs initially exceeds that from gas accretion and remains comparable over the first $\sim200$ Myr. Including disk-induced TDEs shifts the black hole population toward higher masses, increases the abundance of massive black holes at $z\sim9$--10, and produces larger black hole-to-stellar mass ratios. This channel helps alleviate, but does not fully remove, the tension between heavy-seed models and the most extreme high-redshift black hole candidates, suggesting that additional growth mechanisms may still be required.

astro-ph.GA↗

arXiv2026

Rapid Growth of Intermediate-Mass Black Holes through Disk-induced Stellar Disruptions

Dense nuclear star clusters provide unique environments for studying the dynamical interactions between stars and massive black holes. When an accretion disk is present, dissipative star--disk interactions can capture surrounding stars, drive their inward migration, and ultimately lead to disk-induced tidal disruption events\,(dTDEs). The long-term feeding rate from this process, however, cannot be inferred from single-orbit migration estimates alone, as it depends on the coupled evolution of disk capture, collisional relaxation, stellar depletion and replenishment, and physical mergers within the star cluster. In this work, we use high-performance direct $N$-body simulations combined with analytic prescriptions for star--disk interactions to follow this coupled evolution for intermediate-mass black holes\,(IMBHs) with accretion disks embedded in dense stellar clusters. The simulations track the formation of the stellar cusp, the capture of stars by repeated disk crossings, their subsequent orbital damping and migration, and their eventual consumption by the central IMBH. We find that dTDEs can sustain stellar mass supply rates of $\sim10^{-3}\,M_\odot \,\mathrm{yr}^{-1}$, which exceeds the Eddington-limited gas accretion rate for IMBHs with $M_\bullet<10^5\,M_\odot$. These results identify dTDEs as an efficient stellar feeding channel for IMBHs in gas-rich dense stellar systems. As one possible application, this mechanism may help transform $\sim10^3\,M_\odot$ IMBHs into more massive black-hole seeds, provided that compact stellar clusters and accretion disks persist for $>30$ Myr.

astro-ph.GA↗

arXiv2026

Local Law and Outlier Eigenvalues of Spiked Separable Covariance Matrices

We prove local laws for the resolvents of separable covariance matrices of the form $\mathcal Q=A^{1/2}XBX^*A^{1/2}$, where $X=(x_{ij})$ is a $p\times n$ random matrix whose entries $x_{ij}$ are i.i.d.~random variables with mean 0 and variance $n^{-1}$, and $A,B$ are deterministic non-negative definite symmetric (or Hermitian) matrices. Following the method developed in arXiv:1611.05364, we first establish a self-consistent equation for the resolvent of $\mathcal Q$ and use it to prove optimal local laws without the technical assumption $\mathbb{E}[x_{ij}^{3}]=0$, which was essential in the previous derivation of the local laws in arXiv:1809.04572. As an application of our local law, we compute the asymptotic distribution of the outlier eigenvalues for spiked separable covariance matrices, extending the corresponding result in arXiv:2008.11903.

math.PR↗

arXiv2026

OmniInfer: System-Wide Acceleration Techniques for Optimizing LLM Serving Throughput and Latency

Large Language Models drive a wide range of modern AI applications but impose substantial challenges on large-scale serving systems due to intensive computation, strict latency constraints, and throughput bottlenecks. We introduce OmniInfer, a unified system-level acceleration framework designed to maximize end-to-end serving efficiency through fine-grained optimization of expert placement, cache compression, and scheduling. OmniInfer integrates three complementary components: OmniPlacement for load-aware Mixture-of-Experts scheduling, OmniAttn for sparse attention acceleration, and OmniProxy for disaggregation-aware request scheduling. Built atop vLLM, OmniInfer delivers system-wide performance gains through adaptive resource disaggregation, efficient sparsity exploitation, and global coordination across prefill and decode phases. Evaluated on DeepSeek-R1 within a 10-node Ascend 910C cluster, OmniInfer achieves 616 QPM, where the unified framework reduces TPOT by 36\%, and the superimposition of OmniProxy further slashes TTFT by 38\%. The project is open-sourced at [this https URL](https://gitee.com/omniai/omniinfer).

cs.DC↗

arXiv2026

Beyond Model Base Retrieval: Weaving Knowledge to Master Fine-grained Neural Network Design

Designing high-performance neural networks for new tasks requires balancing optimization quality with search efficiency. Current methods fail to achieve this balance: neural architectural search is computationally expensive, while model retrieval often yields suboptimal static checkpoints. To resolve this dilemma, we model the performance gains induced by fine-grained architectural modifications as edit-effect evidence and build evidence graphs from prior tasks. By constructing a retrieval-augmented model refinement framework, our proposed M-DESIGN dynamically weaves historical evidence to discover near-optimal modification paths. M-DESIGN features an adaptive retrieval mechanism that quickly calibrates the evolving transferability of edit-effect evidence from different sources. To handle out-of-distribution shifts, we introduce predictive task planners that extrapolate gains from multi-hop evidence, thereby reducing reliance on an exhaustive repository. Based on our model knowledge base of 67,760 graph neural networks across 22 datasets, extensive experiments demonstrate that M-DESIGN consistently outperforms baselines, achieving the search-space best performance in 26 out of 33 cases under a strict budget.

cs.LG↗

arXiv2026

Single-frame super-resolution via Sparse Point Optimization

Fluorescence microscopy is essential in biological and medical research, providing critical insights into cellular structures. However, limited by optical diffraction and background noise, a substantial amount of hidden information is still unexploited. To address these challenges, we introduce a novel computational method, termed Sparse Point Optimization Theory (SPOT), which accurately localizes fluorescent emitters by solving an optimization problem. Our results demonstrate that SPOT successfully resolves 30 nm fluorescent line pairs, reveals structural details beyond the diffraction limit in both Airyscan and structured illumination microscopy, and outperforms established algorithms in single-molecule localization tasks. This generic method effectively pushes the resolution limit in the presence of noise, and holds great promise for advancing fluorescence microscopy and analysis in cell biology.

physics.bio-ph↗

arXiv2026

Proficient Graph Neural Network Design by Accumulating Knowledge on Large Language Models

High-level automation is increasingly critical in AI, driven by rapid advances in large language models (LLMs) and AI agents. However, LLMs, despite their general reasoning power, struggle significantly in specialized, data-sensitive tasks such as designing Graph Neural Networks (GNNs). This difficulty arises from (1) the inherent knowledge gaps in modeling the intricate, varying relationships between graph properties and suitable architectures and (2) the external noise from misleading descriptive inputs, often resulting in generic or even misleading model suggestions. Achieving proficiency in designing data-aware models -- defined as the meta-level capability to systematically accumulate, interpret, and apply data-specific design knowledge -- remains challenging for existing automated approaches, due to their inefficient construction and application of meta-knowledge. To achieve meta-level proficiency, we propose DesiGNN, a knowledge-centered framework that systematically converts past model design experience into structured, fine-grained knowledge priors well-suited for meta-learning with LLMs. To account for the inherent variability and external noise, DesiGNN aligns empirical property filtering from extensive benchmarks with adaptive elicitation of literature insights via LLMs. By constructing a solid meta-knowledge between unseen graph understanding and known effective architecture patterns, DesiGNN can deliver top-5.77% initial model proposals for unseen datasets within seconds and achieve consistently superior performance with minimal search cost compared to baselines.

cs.LG↗

arXiv2025

Understanding the Embedding Models on Hyper-relational Knowledge Graph

Recently, Hyper-relational Knowledge Graphs (HKGs) have been proposed as an extension of traditional Knowledge Graphs (KGs) to better represent real-world facts with additional qualifiers. As a result, researchers have attempted to adapt classical Knowledge Graph Embedding (KGE) models for HKGs by designing extra qualifier processing modules. However, it remains unclear whether the superior performance of Hyper-relational KGE (HKGE) models arises from their base KGE model or the specially designed extension module. Hence, in this paper, we data-wise convert HKGs to KG format using three decomposition methods and then evaluate the performance of several classical KGE models on HKGs. Our results show that some KGE models achieve performance comparable to that of HKGE models. Upon further analysis, we find that the decomposition methods alter the original HKG topology and fail to fully preserve HKG information. Moreover, we observe that current HKGE models are either insufficient in capturing the graph's long-range dependency or struggle to integrate main-triple and qualifier information due to the information compression issue. To further justify our findings and offer a potential direction for future HKGE research, we propose the FormerGNN framework. This framework employs a qualifier integrator to preserve the original HKG topology, and a GNN-based graph encoder to capture the graph's long-range dependencies, followed by an improved approach for integrating main-triple and qualifier information to mitigate compression issues. Our experimental results demonstrate that FormerGNN outperforms existing HKGE models.

cs.LG↗

arXiv2024

Advancing global aerosol forecasting with artificial intelligence

Aerosol forecasting is essential for air quality warnings, health risk assessment, and climate change mitigation. However, it is more complex than weather forecasting due to the intricate interactions between aerosol physicochemical processes and atmospheric dynamics, resulting in significant uncertainty and high computational costs. Here, we develop an artificial intelligence-driven global aerosol-meteorology forecasting system (AI-GAMFS), which provides reliable 5-day, 3-hourly forecasts of aerosol optical components and surface concentrations at a 0.5° x 0.625° resolution. AI-GAMFS combines Vision Transformer and U-Net in a backbone network, robustly capturing the complex aerosol-meteorology interactions via global attention and spatiotemporal encoding. Trained on 42 years of advanced aerosol reanalysis data and initialized with GEOS Forward Processing (GEOS-FP) analyses, AI-GAMFS delivers operational 5-day forecasts in one minute. It outperforms the Copernicus Atmosphere Monitoring Service (CAMS) global forecasting system, GEOS-FP forecasts, and several regional dust forecasting systems in forecasting most aerosol variables including aerosol optical depth and dust components. Our results mark a significant step forward in leveraging AI to refine physics-based aerosol forecasting, facilitating more accurate global warnings for aerosol pollution events, such as dust storms and wildfires.

physics.ao-ph↗

arXiv2024

Search to Fine-tune Pre-trained Graph Neural Networks for Graph-level Tasks

Recently, graph neural networks (GNNs) have shown its unprecedented success in many graph-related tasks. However, GNNs face the label scarcity issue as other neural networks do. Thus, recent efforts try to pre-train GNNs on a large-scale unlabeled graph and adapt the knowledge from the unlabeled graph to the target downstream task. The adaptation is generally achieved by fine-tuning the pre-trained GNNs with a limited number of labeled data. Despite the importance of fine-tuning, current GNNs pre-training works often ignore designing a good fine-tuning strategy to better leverage transferred knowledge and improve the performance on downstream tasks. Only few works start to investigate a better fine-tuning strategy for pre-trained GNNs. But their designs either have strong assumptions or overlook the data-aware issue for various downstream datasets. Therefore, we aim to design a better fine-tuning strategy for pre-trained GNNs to improve the model performance in this paper. Given a pre-trained GNN, we propose to search to fine-tune pre-trained graph neural networks for graph-level tasks (S2PGNN), which adaptively design a suitable fine-tuning framework for the given labeled data on the downstream task. To ensure the improvement brought by searching fine-tuning strategy, we carefully summarize a proper search space of fine-tuning framework that is suitable for GNNs. The empirical studies show that S2PGNN can be implemented on the top of 10 famous pre-trained GNNs and consistently improve their performance. Besides, S2PGNN achieves better performance than existing fine-tuning strategies within and outside the GNN area. Our code is publicly available at \url{https://anonymous.4open.science/r/code_icde2024-A9CB/}.

cs.LG↗

arXiv2021

AutoGEL: An Automated Graph Neural Network with Explicit Link Information

Recently, Graph Neural Networks (GNNs) have gained popularity in a variety of real-world scenarios. Despite the great success, the architecture design of GNNs heavily relies on manual labor. Thus, automated graph neural network (AutoGNN) has attracted interest and attention from the research community, which makes significant performance improvements in recent years. However, existing AutoGNN works mainly adopt an implicit way to model and leverage the link information in the graphs, which is not well regularized to the link prediction task on graphs, and limits the performance of AutoGNN for other graph tasks. In this paper, we present a novel AutoGNN work that explicitly models the link information, abbreviated to AutoGEL. In such a way, AutoGEL can handle the link prediction task and improve the performance of AutoGNNs on the node classification and graph classification task. Specifically, AutoGEL proposes a novel search space containing various design dimensions at both intra-layer and inter-layer designs and adopts a more robust differentiable search algorithm to further improve efficiency and effectiveness. Experimental results on benchmark data sets demonstrate the superiority of AutoGEL on several tasks.

cs.LG↗

arXiv2015

Preliminary Research on Dual-Energy X-Ray Phase-Contrast Imaging

Dual-energy X-ray absorptiometry (DEXA) has been widely applied to measure bone mineral density (BMD) and soft-tissue composition of human body. However, the use of DEXA is greatly limited for low-Z materials such as soft tissues due to their weak absorption. While X-ray phase-contrast imaging (XPCI) shows significantly improved contrast in comparison with the conventional standard absorption-based X-ray imaging for soft tissues. In this paper, we propose a novel X-ray phase-contrast method to measure the area density of low-Z materials, including a single-energy method and a dual-energy method. The single-energy method is for the area density calculation of one low-Z material, while the dual-energy method is aiming to calculate the area densities of two low-Z materials simultaneously. Comparing the experimental and simulation results with the theoretic ones, the new method proves to have the potential to replace DEXA in area density measurement. The new method sets the prerequisites for future precise and low-dose area density calculation method of low-Z materials.

physics.med-ph↗

arXiv2015

Performance Analysis of Quantitative Phase Retrieval Method in Zernike Phase Contrast X-ray Microscopy

Since the invention of Zernike phase contrast method in 1930, it has been widely used in optical microscopy and more recently in X-ray microscopy. Considering the image contrast is a mixture of absorption and phase information, we recently have proposed and demonstrated a method for quantitative phase retrieval in Zernike phase contrast X-ray microscopy. In this contribution, we analyzed the performance of this method at different photon energies. Intensity images of PMMA samples are simulated at 2.5 keV and 6.2 keV, respectively, and phase retrieval is performed using the proposed method. The results demonstrated that the proposed phase retrieval method is applicable over a wide energy range. For weakly absorbing features, the optimal photon energy is 2.5 keV, from the point of view of image contrast and accuracy of phase retrieval. On the other hand, in the case of strong absorption objects, a higher photon energy is preferred to reduce the error of phase retrieval. Those results can be used as guidelines to perform quantitative phase retrieval in Zernike phase contrast X-ray microscopy with the proposed method.

physics.optics↗

arXiv2015

Research on the background correction method in x-ray phase contrast imaging with Talbot-Lau interferometer

X-ray Talbot-Lau interferometer has been used widely to conduct phase contrast imaging with a conventional low-brilliance x-ray source. Typically, in this technique, background correction has to be performed in order to obtain the pure signal of the sample under inspection. In this study, we reported on a research on the background correction strategies within this technique, especially we introduced a new phase unwrapping solution for one conventional background correction method, the key point of this new solution is changing the initial phase of each pixel by a cyclic shift operation on the raw images collected in phase stepping scan. Experimental result and numerical analysis showed that the new phase unwrapping algorithm could successfully subtract contribution of the system's background without error. Moreover, a potential advantage of this phase unwrapping strategy is that its effective phase measuring range could be tuned flexibly in some degree for example to be (-pi+3, pi+3], thus it would find usage in certain case because measuring range of the currently widely used background correction method is fixed to be (-pi, pi].

physics.optics↗

arXiv2014

Experimental research on the feature of an X-ray Talbot-Lau interferometer vs. tube accelerating voltage

X-ray Talbot-Lau interferometer has been used most widely to perform X-ray phase-contrast imaging with a conventional low-brilliance X-ray source, it yields high-sensitivity phase and dark-field images of sample producing low absorption contrast, thus bearing tremendous potential for future clinical diagnosis. In this manuscript, while changing accelerating voltage of the X-ray tube from 35KV to 45KV, X-ray phase-contrast imaging of a test sample were performed at each integer KV position to investigate the characteristic of an X-ray Talbot-Lau interferometer (located in the Institute of Multidisciplinary Research for Advanced Materials, Tohoku University, Japan) vs. tube voltage. Experimental results and data analysis show that this X-ray Talbot-Lau interferometer is insensitive to the tube accelerating voltage within a certain range, fringe visibility around 44% is maintained in the aforementioned tube voltage range. This experimental research implies that potential new dual energy phase-contrast X-ray imaging strategy and rough refraction spectrum measurement is feasible with this X-ray Talbot-Lau interferometer.

physics.optics↗

arXiv2014

A LabVIEW based user-friendly X-ray phase-contrast imaging system software platform

X-ray phase-contrast imaging can provide greatly improved contrast over conventional absorption-based imaging for weakly absorbing samples, such as biological soft tissues and fibre composites. In this manuscript, we introduce an easy and fast way to develop a user-friendly software platform dedicated to the new grating-based X-ray phase-contrast imaging setup recently built at the National Synchrotron Radiation Laboratory of the University of Science and Technology of China. Unified management and control of 21 motorized positioning stages, of an ultra-precision piezoelectric translation stage and of the X-ray tube are achieved with this platform. The software package also covers the automatic image acquisition of the phase-stepping scanning with a flat panel detector. Moreover, a data post-processing module for signals retrieval and other custom features are in principle available. With a seamless integration of all necessary functions in a unique package, this software platform will greatly support the user activity during experimental runs.

cs.SE↗

arXiv2014

A user-friendly nano-CT image alignment and 3D reconstruction platform based on LabVIEW

X-ray computed tomography at the nanometer scale (nano-CT) offers a wide range of applications in scientific and industrial areas. Here we describe a reliable, user-friendly and fast software package based on LabVIEW that may allow to perform all procedures after the acquisition of raw projection images in order to obtain the inner structure of the investigated sample. A suitable image alignment process to address misalignment problems among image series due to mechanical manufacturing errors, thermal expansion and other external factors has been considered together with a novel fast parallel beam 3D reconstruction procedure, developed ad hoc to perform the tomographic reconstruction. Remarkably improved reconstruction results obtained at the Beijing Synchrotron Radiation Facility after the image calibration confirmed the fundamental role of this image alignment procedure that minimizes unwanted blurs and additional streaking artifacts always present in reconstructed slices. Moreover, this nano-CT image alignment and its associated 3D reconstruction procedure fully based on LabVIEW routines, significantly reduce the data post-processing cycle, thus making faster and easier the activity of the users during experimental runs.

cs.CE↗