arXiv subjects

Wei Zhu

Publications and source records attributed to Wei Zhu.

At least 91 records · Page 5Linked to original sources

arXiv2025

Multi-Branch DNN and CRLB-Ratio-Weight Fusion for Enhanced DOA Sensing via a Massive H$^2$AD MIMO Receiver

As a green MIMO structure, massive H$^2$AD is viewed as a potential technology for the future 6G wireless network. For such a structure, it is a challenging task to design a low-complexity and high-performance fusion of target direction values sensed by different sub-array groups with fewer use of prior knowledge. To address this issue, a lightweight Cramer-Rao lower bound (CRLB)-ratio-weight fusion (WF) method is proposed, which approximates inverse CRLB of each subarray using antenna number reciprocals to eliminate real-time CRLB computation. This reduces complexity and prior knowledge dependence while preserving fusion performance. Moreover, a multi-branch deep neural network (MBDNN) is constructed to further enhance direction-of-arrival (DOA) sensing by leveraging candidate angles from multiple subarrays. The subarray-specific branch networks are integrated with a shared regression module to effectively eliminate pseudo-solutions and fuse true angles. Simulation results show that the proposed CRLB-ratio-WF method achieves DOA sensing performance comparable to CRLB-based methods, while significantly reducing the reliance on prior knowledge. More notably, the proposed MBDNN has superior performance in low-SNR ranges. At SNR $= -15$ dB, it achieves an order-of-magnitude improvement in estimation accuracy compared to CRLB-ratio-WF method.

eess.SP↗

arXiv2025

Robust Moment Identification for Nonlinear PDEs via a Neural ODE Approach

We propose a data-driven framework for learning reduced-order moment dynamics from PDE-governed systems using Neural ODEs. In contrast to derivative-based methods like SINDy, which necessitate densely sampled data and are sensitive to noise, our approach based on Neural ODEs directly models moment trajectories, enabling robust learning from sparse and potentially irregular time series. Using as an application platform the nonlinear Schrödinger equation, the framework accurately recovers governing moment dynamics when closure is available, even with limited and irregular observations. For systems without analytical closure, we introduce a data-driven coordinate transformation strategy based on Stiefel manifold optimization, enabling the discovery of low-dimensional representations in which the moment dynamics become closed, facilitating interpretable and reliable modeling. We also explore cases where a closure model is not known, such as a Fisher-KPP reaction-diffusion system. Here we demonstrate that Neural ODEs can still effectively approximate the unclosed moment dynamics and achieve superior extrapolation accuracy compared to physical-expert-derived ODE models. This advantage remains robust even under sparse and irregular sampling, highlighting the method's robustness in data-limited settings. Our results highlight the Neural ODE framework as a powerful and flexible tool for learning interpretable, low-dimensional moment dynamics in complex PDE-governed systems.

nlin.PS↗

arXiv2025

MOA-2022-BLG-091Lb and KMT-2024-BLG-1209Lb: Microlensing planets detected through weak caustic-crossing signals

The light curves of the microlensing events MOA-2022-BLG-091 and KMT-2024-BLG-1209 exhibit anomalies with very similar features. These anomalies appear near the peaks of the light curves, where the magnifications are moderately high, and are distinguished by weak caustic-crossing features with minimal distortion while the source remains inside the caustic. To achieve a deeper understanding of these anomalies, we conducted a comprehensive analysis of the lensing events. We carried out binary-lens modeling with a thorough exploration of the parameter space. This analysis revealed that the anomalies in both events are of planetary origin, although their exact interpretation is complicated by different types of degeneracy. In the case of MOA-2022-BLG-091, the main difficulty in the interpretation of the anomaly arises from a newly identified degeneracy related to the uncertain angle at which the source trajectory intersects the planet-host axis. For KMT-2024-BLG-1209, the interpretation is affected by the previously known inner-outer degeneracy, which leads to ambiguity between solutions in which the source passes through either the inner or outer caustic region relative to the planet host. Bayesian analysis indicates that the planets in both lens systems are giant planets with masses about 2 to 4 times that of Jupiter, orbiting early K-type main-sequence stars. Both systems are likely located in the Galactic disk at a distance of around 4 kiloparsecs. The degeneracy in KMT-2024-BLG-1209 is challenging to resolve because it stems from intrinsic similarities in the caustic structures of the degenerate solutions. In contrast, the degeneracy in MOA-2022-BLG-091, which occurs by chance rather than from inherent characteristics, is expected to be resolved by the future space based Roman RGES microlensing survey.

astro-ph.EP↗

arXiv2025

KMT-2022-BLG-1818Lb,c: A Cold Super-Jupiter with a Saturn Sibling

We present the discovery and analysis of the sixth microlensing two-planet system, KMT-2022-BLG-1818Lb,c, detected by a follow-up program targeting high-magnification events. Both planets are subject to the well-known ''Close/Wide'' degeneracy, although for the first planet, which has a super-Jovian mass ratio of $q_2 \simeq 5\times 10^{-3}$ in both solutions, the Close topology, with a normalized separation of $s\simeq 0.70$, is clearly preferred by $Δχ^2=26$. However, contrary to all previous two-planet microlensing systems, the mass ratio for the second planet, $q_3$, is substantially (factor of $\sim 10$) different for the Close and Wide topologies of the first planet. While this degeneracy is resolved in the present case due to high-cadence follow-up observations, the appearance of this new degeneracy indicates the need for caution in the analysis of future two-planet systems. A Bayesian analysis suggests that the host is likely a K-dwarf star in the Galactic disk. The first planet is probably a super-Jupiter on a Jupiter-like orbit, while the second planet is a Saturn-class planet on either a Mercury-like or Saturn-like orbit.

astro-ph.EP↗

arXiv2025

Microlensing events indicate that super-Earth exoplanets are common in Jupiter-like orbits

Exoplanets classified as super-Earths are commonly observed on short period orbits, close to their host stars, but their abundance on wider orbits is poorly constrained. Gravitational microlensing is sensitive to exoplanets on wide orbits. We observed the microlensing event OGLE-2016-BLG-0007, which indicates an exoplanet with a planet-to-star mass ratio roughly double the Earth-Sun mass-ratio, on an orbit longer than Saturn's. We combine this event with a larger sample from a microlensing survey to determine the distribution of mass ratios for planets on wide orbits. We infer there are $\sim 0.35$ super-Earth planets per star on Jupiter-like orbits. The observations are most consistent with a bimodal distribution, with separate peaks for super-Earths and gas giants. We suggest that this reflects differences in their formation processes.

astro-ph.EP↗

arXiv2025

Systematic KMTNet Planetary Anomaly Search. XII. Complete Sample of 2017 Subprime Field Planets

We report the analysis of four unambiguous planets and one possible planet from the subprime fields ($Γ\leq 1~{\rm hr}^{-1}$) of the 2017 Korea Microlensing Telescope Network (KMTNet) microlensing survey, to complete the KMTNet AnomalyFinder planetary sample for the 2017 subprime fields. They are KMT-2017-BLG-0849, KMT-2017-BLG-1057, OGLE-2017-BLG-0364, and KMT-2017-BLG-2331 (unambiguous), as well as KMT-2017-BLG-0958 (possible). For the four unambiguous planets, the mean planet-host mass ratios, $q$, are $(1.0, 1.2, 4.6, 13) \times 10^{-4}$, the median planetary masses are $(6.4, 24, 76, 171)~M_{\oplus}$ and the median host masses are $(0.19, 0.57, 0.49, 0.40)~M_{\odot}$ from a Bayesian analysis. We have completed the AnomalyFinder planetary sample from the first 4-year KMTNet data (2016--2019), with 112 unambiguous planets in total, which nearly tripled the microlensing planetary sample. The ``sub-Saturn desert'' ($\log q = \left[-3.6, -3.0\right]$) found in the 2018 and 2019 KMTNet samples is confirmed by the 2016 and 2017 KMTNet samples.

astro-ph.EP↗

arXiv2025

MOA-2020-BLG-135Lb: A New Neptune-class Planet for the Extended MOA-II Exoplanet Microlens Statistical Analysis

We report the light-curve analysis for the event MOA-2020-BLG-135, which leads to the discovery of a new Neptune-class planet, MOA-2020-BLG-135Lb. With a derived mass ratio of $q=1.52_{-0.31}^{+0.39} \times 10^{-4}$ and separation $s\approx1$, the planet lies exactly at the break and likely peak of the exoplanet mass-ratio function derived by the MOA collaboration (Suzuki et al. 2016). We estimate the properties of the lens system based on a Galactic model and considering two different Bayesian priors: one assuming that all stars have an equal planet-hosting probability and the other that planets are more likely to orbit more massive stars. With a uniform host mass prior, we predict that the lens system is likely to be a planet of mass $m_\mathrm{planet}=11.3_{-6.9}^{+19.2} M_\oplus$ and a host star of mass $M_\mathrm{host}=0.23_{-0.14}^{+0.39} M_\odot$, located at a distance $D_L=7.9_{-1.0}^{+1.0}\;\mathrm{kpc}$. With a prior that holds that planet occurrence scales in proportion to the host star mass, the estimated lens system properties are $m_\mathrm{planet}=25_{-15}^{+22} M_\oplus$, $M_\mathrm{host}=0.53_{-0.32}^{+0.42} M_\odot$, and $D_L=8.3_{-1.0}^{+0.9}\; \mathrm{kpc}$. This planet qualifies for inclusion in the extended MOA-II exoplanet microlens sample.

astro-ph.EP↗

arXiv2025

Dependence of Planet populations on Stellar Mass and Metallicity: A Pebble Accretion-based Planet Population Synthesis

The formation and evolution of planetary systems are linked to their host stellar environment. In this study, we employ a pebble accretion-based planet population synthesis model to explore the correlation between planetary properties and stellar mass/metallicity. Our numerical results reproduce several main aspects of exoplanetary observations. First, we find that the occurrence rate of super-Earths $η_{\rm SE}$ follows an inverted V-shape in relation to stellar mass: it increases with stellar mass among lower-mass dwarfs, peaks at early-M dwarfs, and declines toward higher-mass GK stars. Second, super-Earths grow ubiquitously around stars with various metallicities, exhibiting a flat or weak $η_{\rm SE}$ dependence on $Z_{\star}$. Third, giant planets, in contrast, form more frequently around stars with higher-mass/metallicity. Lastly, we extend a subset of simulations to $1$ Gyr to investigate the long-term evolution of the systems' architecture. By converting our simulated systems into synthetic observations, we find that the eccentricities and inclinations of single-transit systems increase with stellar metallicity, while these dependencies in multi-planet systems remains relatively weak. The alignment between our results and observations provides key insights into the connection between planet populations and stellar properties.

astro-ph.EP↗

arXiv2025

ML-Triton, A Multi-Level Compilation and Language Extension to Triton GPU Programming

In the era of LLMs, dense operations such as GEMM and MHA are critical components. These operations are well-suited for parallel execution using a tilebased approach. While traditional GPU programming often relies on low level interfaces like CUDA or SYCL, Triton has emerged as a DSL that offers a more user-friendly and portable alternative by programming at a higher level. The current Triton starts at the workgroup (aka threadblock) level, and directly lowers to per-thread level. And then attempt to coalesce and amend through a series of passes, promoting information from low-level representation. We believe this is pre-mature lowering based on the below observations. 1. GPU has a hierarchical structure both physically and logically. Modern GPUs often feature SIMD units capable of directly operating on tiles on a warp or warpgroup basis, such as blocked load and blocked MMA. 2. Multi-level gradual lowering can make compiler decoupled and clean by separating considerations inter and intra a logical layer. 3. Kernel developers often need fine control to get good performance on the latest hardware. FlashAttention2 advocates explicit data partition between warps to make a performance boost. In this context, we propose ML-Triton which features multi-level compilation flow and programming interface. Our approach begins at the workgroup level and progressively lowers to the warp and intrinsic level, implementing a multilevel lowering align with the hierarchical nature of GPU. Additionally, we extend triton language to support user-set compiler hint and warp level programming, enabling researchers to get good out-of-the box performance without awaiting compiler updates. Experimental results demonstrate that our approach achieves performance above 95% of expert-written kernels on Intel GPU, as measured by the geometric mean.

cs.CL↗

arXiv2025

Image Restoration Models with Optimal Transport and Total Variation Regularization

In this paper, we propose image restoration models using optimal transport (OT) and total variation regularization. We present theoretical results of the proposed models based on the relations between the dual Lipschitz norm from OT and the G-norm introduced by Yves Meyer. We design a numerical method based on the Primal-Dual Hybrid Gradient (PDHG) algorithm for the Wasserstain distance and the augmented Lagrangian method (ALM) for the total variation, and the convergence analysis of the proposed numerical method is established. We also consider replacing the total variation in our model by one of its modifications developed in \cite{zhu}, with the aim of suppressing the stair-casing effect and preserving image contrasts. Numerical experiments demonstrate the features of the proposed models.

math.NA↗

arXiv2025

Statistical Guarantees of Group-Invariant GANs

This work presents the first statistical performance guarantees for group-invariant generative models. Many real data, such as images and molecules, are invariant to certain group symmetries, which can be taken advantage of to learn more efficiently as we rigorously demonstrate in this work. Here we specifically study generative adversarial networks (GANs), and quantify the gains when incorporating symmetries into the model. Group-invariant GANs are a type of GANs in which the generators and discriminators are hardwired with group symmetries. Empirical studies have shown that these networks are capable of learning group-invariant distributions with significantly improved data efficiency. In this study, we aim to rigorously quantify this improvement by analyzing the reduction in sample complexity and in the discriminator approximation error for group-invariant GANs. Our findings indicate that when learning group-invariant distributions, the number of samples required for group-invariant GANs decreases proportionally by a factor of the group size and the discriminator approximation error has a reduced lower bound. Importantly, the overall error reduction cannot be achieved merely through data augmentation on the training data. Numerical results substantiate our theory and highlight the stark contrast between learning with group-invariant GANs and using data augmentation. This work also sheds light on the study of other generative models with group symmetries, such as score-based generative models.

stat.ML↗

arXiv2025

SMTPD: A New Benchmark for Temporal Prediction of Social Media Popularity

Social media popularity prediction task aims to predict the popularity of posts on social media platforms, which has a positive driving effect on application scenarios such as content optimization, digital marketing and online advertising. Though many studies have made significant progress, few of them pay much attention to the integration between popularity prediction with temporal alignment. In this paper, with exploring YouTube's multilingual and multi-modal content, we construct a new social media temporal popularity prediction benchmark, namely SMTPD, and suggest a baseline framework for temporal popularity prediction. Through data analysis and experiments, we verify that temporal alignment and early popularity play crucial roles in social media popularity prediction for not only deepening the understanding of temporal dynamics of popularity in social media but also offering a suggestion about developing more effective prediction models in this field. Code is available at https://github.com/zhuwei321/SMTPD.

cs.SI↗

arXiv2025

Computer Assisted Discovery of Integrability via SILO: Sparse Identification of Lax Operators

We formulate the discovery of Lax integrability of Hamiltonian dynamical systems as a symbolic regression problem, which, loosely speaking, seeks to maximize the compatibility between a pair of Lax operators and the known Hamiltonian of the dynamical system. Our approach is first tested on the simple harmonic oscillator. We then move on to the Henon-Heiles system, i.e. a two-degree-of-freedom system of nonlinear oscillators. The integrability of the Henon-Heiles system is critically dependent on a set of three parameters within its Hamiltonian, a fact that we leverage to assess the robustness of our approach in detecting the integrability of this system with respect to the parameter dependence of the Hamiltonian. We then adapt our method to canonical examples of Hamiltonian partial differential equations, including the Korteweg-de Vries and cubic nonlinear Schrödinger equations, again testing robustness against nonintegrable perturbations of their respective Hamiltonians. In all examples, our approach reliably confirms or denies the integrability of the equations of interest. Moreover, by appropriately adjusting the loss function and applying thresholded $l^0$ regularization to enforce sparsity in the operator weights, we successfully recover accurate forms of the Lax pairs despite wide initial hypotheses on the operators. Some of the relevant Lax pairs, notably for the Henon-Heiles system and the Korteweg-deVries equation, are distinct from the ones that are typically reported in the literature. The Lax pairs that our methodology discovers warrant further mathematical and computational investigation, and we discuss extensively the opportunities for further improvement of SILO as a viable tool for interpretable exploration of integrable Hamiltonian dynamical systems.

nlin.SI↗

arXiv2025

Continuously tunable anomalous Hall crystals in rhombohedral heptalayer graphene

The interplay of electronic interactions and nontrivial topology can give rise to a wealth of exotic quantum states. A notable example is the formation of Wigner crystals driven by strong electron-electron interactions. When these electronic crystals emerge in a parent band carrying a large Berry curvature, they can exhibit topologically nontrivial properties as anomalous Hall crystals, spontaneously breaking both continuous translational symmetry and time-reversal symmetry. Here, we report the experimental observation of tunable anomalous Hall crystals in rhombohedral heptalayer graphene moiré superlattices. At filling factors near one electron per moiré unit cell (v=1), we identify a series of incommensurate Chern insulators with a Chern number of C=1. Furthermore, we observe spontaneous time-reversal symmetry breaking spanning the entire filling range from v=1 to v=2, manifesting as anomalous Hall effects with pronounced magnetic hysteresis. Notably, anomalous Hall crystals with a high Chern number C=3 are observed over generic fillings ranging from v=1.5 to v=2. These anomalous Hall crystals are incommensurate with the moiré superlattice and exhibit dispersive fan diagrams consistent with the Streda formula, with their positions continuously tunable through displacement fields. Remarkably, these partially filled Chern insulators display Chern numbers distinct from their parent bands. Our findings demonstrate the rich variety of electronic crystalline states in rhombohedral graphene moiré superlattices, offering valuable insights into the strongly correlated topological phases.

cond-mat.mes-hall↗

arXiv2025

Robust non-Abelian even-denominator fractional Chern insulator in twisted bilayer MoTe$_2$

A recent experiment observes a series of quantum spin Hall effects in transition metal dichalcogenide moiré MoTe$_2$ [K. Kang, et. al, Nature 628, 522-526 (2024)]. Among them, the vanishing Hall signal at the filling factor $ν=3$ implies a possible realization of a time-reversal pair of the even-denominator fractional Chern insulators (FCIs). Inspired by this discovery, we investigate whether a robust incompressible quantum Hall liquid can be stabilized in the half-filled Chern band of twisted MoTe$_2$ bilayers. We use the continuum model with parameters relevant to twisted MoTe$_2$ bilayers and obtain three consecutive nearly flat Chern bands with the same Chern number. Crucially, when the second moiré miniband is half-filled,signatures of non-Abelian frctional quantum Hall state are found via exact diagonalization calculations, including the stable six-fold ground state degeneracy that grows more robust with the lattice size and is consistent with an even-denominator FCI state. We further perform flux insertion simulations to reveal a 1/2 quantized many-body Chern number as direct evidence of topological order. Furthermore, the ground state density structure factors show no sharp peak, indicating no charge density wave order. These evidences signal the potential of realizing the non-Abelian state at zero magnetic field in twisted bilayer MoTe$_2$ at the fractional hole filling 3/2.

cond-mat.str-el↗

arXiv2025

Learning Novel Transformer Architecture for Time-series Forecasting

Despite the success of Transformer-based models in the time-series prediction (TSP) tasks, the existing Transformer architecture still face limitations and the literature lacks comprehensive explorations into alternative architectures. To address these challenges, we propose AutoFormer-TS, a novel framework that leverages a comprehensive search space for Transformer architectures tailored to TSP tasks. Our framework introduces a differentiable neural architecture search (DNAS) method, AB-DARTS, which improves upon existing DNAS approaches by enhancing the identification of optimal operations within the architecture. AutoFormer-TS systematically explores alternative attention mechanisms, activation functions, and encoding operations, moving beyond the traditional Transformer design. Extensive experiments demonstrate that AutoFormer-TS consistently outperforms state-of-the-art baselines across various TSP benchmarks, achieving superior forecasting accuracy while maintaining reasonable training efficiency.

cs.LG↗

arXiv2025

Adapting Large Language Models for Time Series Modeling via a Novel Parameter-efficient Adaptation Method

Time series modeling holds significant importance in many real-world applications and has been extensively studied. While pre-trained foundation models have made impressive strides in the fields of natural language processing (NLP) and computer vision (CV), their development in time series domains has been constrained by data sparsity. A series of recent studies have demonstrated that large language models (LLMs) possess robust pattern recognition and reasoning abilities over complex sequences of tokens. However, the current literature have yet striked a high-quality balance between (a) effectively aligning the time series and natural language modalities, and (b) keeping the inference efficiency. To address the above issues, we now propose the Time-LlaMA framework. Time-LlaMA first converts the time series input into token embeddings through a linear tokenization mechanism. Second, the time series token embeddings are aligned with the text prompts. Third, to further adapt the LLM backbone for time series modeling, we have developed a dynamic low-rank adaptation technique (D-LoRA). D-LoRA dynamically chooses the most suitable LoRA modules at each layer of the Transformer backbone for each time series input, enhancing the model's predictive capabilities. Our experimental results on an extensive collection of challenging real-world time series tasks confirm that our proposed method achieves the state-of-the-art (SOTA) performance.

cs.CL↗

arXiv2025

A differentiable binary microlensing model using adaptive contour integration method

We present microlux, which is a Jax-based code that can compute the binary microlensing light curve and its derivatives both efficiently and accurately. The key feature of microlux is the implementation of a modified version of the adaptive sampling algorithm that was originally proposed by V. Bozza to account for the finite-source effect most efficiently. The efficiency and accuracy of microlux have been verified across the relevant parameter space for binary microlensing. As a differentiable code, microlux makes it possible to apply gradient-based algorithms to the search and posterior estimation of the microlensing modeling. As an example, we use microlux to model a real microlensing event and infer the model posterior via both Fisher information matrix and Hamiltonian Monte Carlo, neither of which would have been possible without the access to accurate model gradients.

astro-ph.IM↗