arXiv subjects

Feng Wang

Publications and source records attributed to Feng Wang.

At least 145 records · Page 8Linked to original sources

arXiv2025

Structure-Adaptive Topology Optimization Framework for Photonic Band Gaps with TM-Polarized Sources

We present a structure-adaptive topology optimization framework for engineering photonic band gaps with TM-polarized sources based on computation of the photonic density of states with a uniform source substituting for the standard Dirac delta function sources in formalisms analogous to $Γ$-point integration and to integration over a full Brillouin zone. We generalize the limiting uniform and Dirac delta function sources to more general collections of sources, such that the union of the sources in a given collection is hyperuniform. The uniform-source approach necessarily leads to the fastest computations. We also demonstrate how our approach can be generalized to the treatment of the frequency-dependent optical response of materials. Finally, we show that we can recover known two-dimensional photonic crystals for the TM polarization. A key advantage of our work is its ability to optimize for a specific midgap frequency and band gap in a structure-adaptive manner. Our work leverages the insight that the determination of the minimum supercell size and the minimum precision to which the frequencies within the photonic band gap must be sampled will lead to the observation of photonic-crystal structures when the $Γ$-point formalism for the uniform-source approach is employed. Additionally, our $Γ$-point and full Brillouin zone formalisms for the uniform-source approach inherently encourage binarized designs even in gradient descent.

physics.optics↗

arXiv2025

BAROC: Concealing Packet Losses in LSNs with Bimodal Behavior Awareness for Livecast Ingestion

The advent of Low-Earth Orbit satellite networks (LSNs), exemplified by initiatives like \emph{Starlink}, \emph{OneWeb} and \emph{Kuiper}, has ushered in a new era of ``Internet from Space" global connectivity. Recent studies have shown that LSNs are capable of providing unprecedented download capacity and low latency to support Livecast viewing. However, Livecast ingestion still faces significant challenges, such as limited uplink capacity, bandwidth degradation, and the burst of packet loss due to frequent satellite reallocations, which cause previous recovery and adaptive solutions to be inferior under this new scenario. In this paper, we conduct an in-depth measurement study dedicated to understanding the implications of satellite reallocations, which reveals that the network status during reallocations with network anomalies exhibits a different distribution, leading to bimodal behaviors on the overall network performance. Motivated by this finding, we propose BAROC, a framework that can effectively conceal burst packet losses by combining a novel proposed MTP-Informer with bimodal behavior awareness during satellite reallocation. BAROC enhances video QoE on the server side by addressing the above challenges and jointly determining the optimal video encoding and recovery parameters. Our extensive evaluation shows that BAROC outperforms other video delivery recovery approaches, achieving an average PSNR improvement of $1.95$ dB and a maximum of $3.44$ dB, along with enhancements in frame rate and parity packet utilization. Additionally, a comprehensive ablation study is conducted to assess the effectiveness of MTP-Informer and components in BAROC.

cs.NI↗

arXiv2025

Decoding the Competing Effects of Dynamic Solvation Structures on Nuclear Magnetic Resonance Chemical Shifts of Battery Electrolytes via Machine Learning

Understanding the solvation structure of electrolytes is critical for optimizing the electrochemical performance of rechargeable batteries, as it directly influences properties such as ionic conductivity, viscosity, and electrochemical stability. The highly complex structures and strong interactions in high-concentration electrolytes make accurate modeling and interpretation of their ``structure-property" relationships even more challenging with spectroscopic methods. In this study, we present a machine learning-based approach to predict dynamic $^7$Li NMR chemical shifts in LiFSI/DME electrolyte solutions. Additionally, we provide a comprehensive structural analysis to interpret the observed chemical shift behavior in our experiments, particularly the abrupt changes in $^7$Li chemical shifts at high concentrations. Using advanced modeling techniques, we quantitatively establish the relationship between molecular structure and NMR spectra, offering critical insights into solvation structure assignments. Our findings reveal the coexistence of two competing local solvation structures that shift in dominance as electrolyte concentration approaches the concentrated limit, leading to anomalous reverse of $^7$Li NMR chemical shift in our experiment. This work provides a detailed molecular-level understanding of the intricate solvation structures probed by NMR spectroscopy, leading the way for enhanced electrolyte design.

physics.chem-ph↗

arXiv2025

Large-Scale Contextual Market Equilibrium Computation through Deep Learning

Market equilibrium is one of the most fundamental solution concepts in economics and social optimization analysis. Existing works on market equilibrium computation primarily focus on settings with relatively few buyers. Motivated by this, our paper investigates the computation of market equilibrium in scenarios with a large-scale buyer population, where buyers and goods are represented by their contexts. Building on this realistic and generalized contextual market model, we introduce MarketFCNet, a deep learning-based method for approximating market equilibrium. We start by parameterizing the allocation of each good to each buyer using a neural network, which depends solely on the context of the buyer and the good. Next, we propose an efficient method to unbiasedly estimate the loss function of the training algorithm, enabling us to optimize the network parameters through gradient. To evaluate the approximated solution, we propose a metric called Nash Gap, which quantifies the deviation of the given allocation and price pair from the market equilibrium. Experimental results indicate that MarketFCNet delivers competitive performance and significantly lower running times compared to existing methods as the market scale expands, demonstrating the potential of deep learning-based methods to accelerate the approximation of large-scale contextual market equilibrium.

cs.GT↗

arXiv2025

Baichuan 2: Open Large-scale Language Models

Large language models (LLMs) have demonstrated remarkable performance on a variety of natural language tasks based on just a few examples of natural language instructions, reducing the need for extensive feature engineering. However, most powerful LLMs are closed-source or limited in their capability for languages other than English. In this technical report, we present Baichuan 2, a series of large-scale multilingual language models containing 7 billion and 13 billion parameters, trained from scratch, on 2.6 trillion tokens. Baichuan 2 matches or outperforms other open-source models of similar size on public benchmarks like MMLU, CMMLU, GSM8K, and HumanEval. Furthermore, Baichuan 2 excels in vertical domains such as medicine and law. We will release all pre-training model checkpoints to benefit the research community in better understanding the training dynamics of Baichuan 2.

cs.CL↗

arXiv2025

Respiratory Differencing: Enhancing Pulmonary Thermal Ablation Evaluation Through Pre- and Intra-Operative Image Fusion

CT image-guided thermal ablation is widely used for lung cancer treatment; however, follow-up data indicate that physicians' subjective assessments of intraoperative images often overestimate the ablation effect, potentially leading to incomplete treatment. To address these challenges, we developed \textit{Respiratory Differencing}, a novel intraoperative CT image assistance system aimed at improving ablation evaluation. The system first segments tumor regions in preoperative CT images and then employs a multi-stage registration process to align these images with corresponding intraoperative or postoperative images, compensating for respiratory deformations and treatment-induced changes. This system provides two key outputs to help physicians evaluate intraoperative ablation. First, differential images are generated by subtracting the registered preoperative images from the intraoperative ones, allowing direct visualization and quantitative comparison of pre- and post-treatment differences. These differential images enable physicians to assess the relative positions of the tumor and ablation zones, even when the tumor is no longer visible in post-ablation images, thus improving the subjective evaluation of ablation effectiveness. Second, the system provides a quantitative metric that measures the discrepancies between the tumor area and the treatment zone, offering a numerical assessment of the overall efficacy of ablation.This pioneering system compensates for complex lung deformations and integrates pre- and intra-operative imaging data, enhancing quality control in cancer ablation treatments. A follow-up study involving 35 clinical cases demonstrated that our system significantly outperforms traditional subjective assessments in identifying under-ablation cases during or immediately after treatment, highlighting its potential to improve clinical decision-making and patient outcomes.

cs.CV↗

arXiv2025

Statistics of velocity gradient and vortex sheet structures in polymeric turbulent von K{á}rm{á}n swirling flow

Investigations into the effects of polymers on small-scale statistics and flow patterns were conducted in a turbulent von Karman swirling (VKS) flow. We employed the tomographic particle image velocimetry (Tomo-PIV) technique to obtain full information on three-dimensional velocity data, allowing us to effectively resolve dissipation scales. Under varying Reynolds numbers ($R_λ=168 - 235$) and polymer concentrations ($ϕ=0 -25~\rm ppm$), we measured the velocity gradient tensor (VGT) and related quantities. Our findings reveal that the ensemble average and probability density function (PDF) of VGT invariants, which represent turbulent dissipation and enstrophy along with their generation terms, are suppressed as polymer concentration increases. Notably, the joint PDFs of the invariants of VGT, which characterize local flow patterns, exhibited significant changes. Specifically, the third-order invariants, especially the local vortex stretching, are greatly suppressed, and strong events of dissipation and enstrophy coexist in space. The local flow pattern tends to be two-dimensional, where the eigenvalues of the rate-of-strain tensor satisfy a ratio $1:0:-1$, and the vorticity aligns with the intermediate eigenvector of the rate-of-strain tensor while is perpendicular to the other two. We find that these statistics observations can be well described by the vortex sheet model. Moreover, we find that these vortex sheet structures align with the symmetry axis of the VKS system and orient randomly in the horizontal plane. Further investigation, including flow visualization and conditional statistics on vorticity, confirms the presence of vortex sheet structures in turbulent flows with polymer additions. Our results establish a link between single-point statistics and small-scale flow topology, shedding light on the previously overlooked small-scale structures in polymeric turbulence.

physics.flu-dyn↗

arXiv2025

Towards Calibrating Financial Market Simulators with High-frequency Data

The fidelity of financial market simulation is restricted by the so-called "non-identifiability" difficulty when calibrating high-frequency data. This paper first analyzes the inherent loss of data information in this difficulty, and proposes to use the Kolmogorov-Smirnov test (K-S) as the objective function for high-frequency calibration. Empirical studies verify that K-S has better identifiability of calibrating high-frequency data, while also leads to a much harder multi-modal landscape in the calibration space. To this end, we propose the adaptive stochastic ranking based negatively correlated search algorithm for improving the balance between exploration and exploitation. Experimental results on both simulated data and real market data demonstrate that the proposed method can obtain up to 36.0% improvement in high-frequency data calibration problems over the compared methods.

cs.CE↗

arXiv2025

LayerAnimate: Layer-level Control for Animation

Traditional animation production decomposes visual elements into discrete layers to enable independent processing for sketching, refining, coloring, and in-betweening. Existing anime generation video methods typically treat animation as a distinct data domain different from real-world videos, lacking fine-grained control at the layer level. To bridge this gap, we introduce LayerAnimate, a novel video diffusion framework with layer-aware architecture that empowers the manipulation of layers through layer-level controls. The development of a layer-aware framework faces a significant data scarcity challenge due to the commercial sensitivity of professional animation assets. To address the limitation, we propose a data curation pipeline featuring Automated Element Segmentation and Motion-based Hierarchical Merging. Through quantitative and qualitative comparisons, and user study, we demonstrate that LayerAnimate outperforms current methods in terms of animation quality, control precision, and usability, making it an effective tool for both professional animators and amateur enthusiasts. This framework opens up new possibilities for layer-level animation applications and creative flexibility. Our code is available at https://layeranimate.github.io.

cs.CV↗

arXiv2025

Commercial Dishes Can Be My Ladder: Sustainable and Collaborative Data Offloading in LEO Satellite Networks

Low Earth Orbit (LEO) satellite networks, characterized by their high data throughput and low latency, have gained significant interest from both industry and academia. Routing data efficiently within these networks is essential for maintaining a high quality of service. However, current routing strategies, such as bent-pipe and inter-satellite link (ISL) routing, have their unique challenges. The bent-pipe strategy requires a dense deployment of dedicated ground stations, while the ISL-based strategy can negatively impact satellite battery lifespan due to increased traffic load, leading to sustainability issues. In this paper, we propose sustainable collaborative offloading, a framework that orchestrates groups of existing commercial resources like ground stations and 5G base stations for data offloading. This orchestration enhances total capacity, overcoming the limitations of a single resource. We propose the collaborator group set construction algorithm to construct candidate groups and the collaborator selection and total payment algorithm to select offloading targets and determine payments no less than the costs. Extensive real-world-based simulations show that our solution significantly improves energy consumption, satellite service life, and end-to-end latency.

cs.NI↗

arXiv2025

Keywords and Instances: A Hierarchical Contrastive Learning Framework Unifying Hybrid Granularities for Text Generation

Contrastive learning has achieved impressive success in generation tasks to militate the "exposure bias" problem and discriminatively exploit the different quality of references. Existing works mostly focus on contrastive learning on the instance-level without discriminating the contribution of each word, while keywords are the gist of the text and dominant the constrained mapping relationships. Hence, in this work, we propose a hierarchical contrastive learning mechanism, which can unify hybrid granularities semantic meaning in the input text. Concretely, we first propose a keyword graph via contrastive correlations of positive-negative pairs to iteratively polish the keyword representations. Then, we construct intra-contrasts within instance-level and keyword-level, where we assume words are sampled nodes from a sentence distribution. Finally, to bridge the gap between independent contrast levels and tackle the common contrast vanishing problem, we propose an inter-contrast mechanism that measures the discrepancy between contrastive keyword nodes respectively to the instance distribution. Experiments demonstrate that our model outperforms competitive baselines on paraphrasing, dialogue generation, and storytelling tasks.

cs.CL↗

arXiv2025

ReaderLM-v2: Small Language Model for HTML to Markdown and JSON

We present ReaderLM-v2, a compact 1.5 billion parameter language model designed for efficient web content extraction. Our model processes documents up to 512K tokens, transforming messy HTML into clean Markdown or JSON formats with high accuracy -- making it an ideal tool for grounding large language models. The model's effectiveness results from two key innovations: (1) a three-stage data synthesis pipeline that generates high quality, diverse training data by iteratively drafting, refining, and critiquing web content extraction; and (2) a unified training framework combining continuous pre-training with multi-objective optimization. Intensive evaluation demonstrates that ReaderLM-v2 outperforms GPT-4o-2024-08-06 and other larger models by 15-20\% on carefully curated benchmarks, particularly excelling at documents exceeding 100K tokens, while maintaining significantly lower computational requirements.

cs.CL↗

arXiv2025

EACO-RAG: Towards Distributed Tiered LLM Deployment using Edge-Assisted and Collaborative RAG with Adaptive Knowledge Update

Large language models (LLMs) have demonstrated impressive capabilities in language tasks, but they require high computing power and rely on static knowledge. To overcome these limitations, Retrieval-Augmented Generation (RAG) incorporates up-to-date external information into LLMs without extensive fine-tuning. Meanwhile, small language models (SLMs) deployed on edge devices offer efficiency and low latency but often struggle with complex reasoning tasks. Unfortunately, current RAG approaches are predominantly based on centralized databases and have not been adapted to address the distinct constraints associated with deploying SLMs in edge environments. To bridge this gap, we propose Edge-Assisted and Collaborative RAG (EACO-RAG), a lightweight framework that leverages distributed edge nodes for adaptive knowledge updates and retrieval. EACO-RAG also employs a hierarchical collaborative gating mechanism to dynamically select among local, edge-assisted, and cloud-based strategies, with a carefully designed algorithm based on Safe Online Bayesian Optimization to maximize the potential performance enhancements. Experimental results demonstrate that EACO-RAG matches the accuracy of cloud-based knowledge graph RAG systems while reducing total costs by up to 84.6% under relaxed delay constraints and by 65.3% under stricter delay requirements. This work represents our initial effort toward achieving a distributed and scalable tiered LLM deployments, with EACO-RAG serving as a promising first step in unlocking the full potential of hybrid edge-cloud intelligence.

cs.DC↗

arXiv2025

UD-Mamba: A pixel-level uncertainty-driven Mamba model for medical image segmentation

Recent advancements have highlighted the Mamba framework, a state-space model known for its efficiency in capturing long-range dependencies with linear computational complexity. While Mamba has shown competitive performance in medical image segmentation, it encounters difficulties in modeling local features due to the sporadic nature of traditional location-based scanning methods and the complex, ambiguous boundaries often present in medical images. To overcome these challenges, we propose Uncertainty-Driven Mamba (UD-Mamba), which redefines the pixel-order scanning process by incorporating channel uncertainty into the scanning mechanism. UD-Mamba introduces two key scanning techniques: 1) sequential scanning, which prioritizes regions with high uncertainty by scanning in a row-by-row fashion, and 2) skip scanning, which processes columns vertically, moving from high-to-low or low-to-high uncertainty at fixed intervals. Sequential scanning efficiently clusters high-uncertainty regions, such as boundaries and foreground objects, to improve segmentation precision, while skip scanning enhances the interaction between background and foreground regions, allowing for timely integration of background information to support more accurate foreground inference. Recognizing the advantages of scanning from certain to uncertain areas, we introduce four learnable parameters to balance the importance of features extracted from different scanning methods. Additionally, a cosine consistency loss is employed to mitigate the drawbacks of transitioning between uncertain and certain regions during the scanning process. Our method demonstrates robust segmentation performance, validated across three distinct medical imaging datasets involving pathology, dermatological lesions, and cardiac tasks.

eess.IV↗

arXiv2025

Competition between excitonic insulators and quantum Hall states in correlated electron-hole bilayers

Excitonic insulators represent a unique quantum phase of matter, providing a rich ground for studying exotic quantum bosonic states. Strongly coupled electron-hole bilayers, which host stable dipolar exciton fluids with an exciton density that can be adjusted electrostatically, offer an ideal platform to investigate correlated excitonic insulators. Based on electron-hole bilayers made of MoSe2/hBN/WSe2 heterostructures, here we study the behavior of excitonic insulators in a perpendicular magnetic field. We report the observation of excitonic quantum oscillations in both Coulomb drag signals and electrical resistance at low to medium magnetic fields. Under a strong magnetic field, we identify multiple quantum phase transitions between the excitonic insulator phase and the bilayer quantum Hall insulator phase. These findings underscore the interplay between the electron-hole interactions and Landau level quantization that opens new possibilities for exploring quantum phenomena in composite bosonic insulators.

cond-mat.mes-hall↗

arXiv2025

A weakly compressible SPH method for RANS simulation of wall-bounded turbulent flows

This paper presents a Weakly Compressible Smoothed Particle Hydrodynamics (WCSPH) method for solving the two-equation Reynolds-Averaged Navier-Stokes (RANS) model. The turbulent wall-bounded flow with or without mild flow separation, a crucial flow pattern in engineering applications, yet rarely explored in the SPH community, is simulated. The inconsistency between the Lagrangian characteristic and RANS model, mainly due to the intense particle shear and near-wall discontinuity, is firstly revealed and addressed by the mainstream and nearwall improvements, respectively. The mainstream improvements, including Adaptive Riemann-eddy Dissipation (ARD) and Limited Transport Velocity Formulation (LTVF), address dissipation incompatibility and turbulent kinetic energy over-prediction issues. The nearwall improvements, such as the particle-based wall model realization, weighted near-wall compensation scheme, and constant $y_p$ strategy, improve the accuracy and stability of the adopted wall model, where the wall dummy particles are still used for future coupling of solid dynamics. Besides, to perform rigorous convergence tests, an level-set-based boundary-offset technique is developed to ensure consistent $y^+$ across different resolutions. The benchmark wall-bounded turbulent cases, including straight, mildly- and strongly-curved, and Half Converging and Diverging (HCD) channels are calculated. Good convergence is, to our best knowledge, firstly achieved for both velocity and turbulent kinetic energy for the SPH-RANS method. All the results agree well with the data from the experiments or simulated by the Eulerian methods at engineering-acceptable resolutions. The proposed method bridges particle-based and mesh-based RANS models, providing adaptability for other turbulence models and potential for turbulent fluid-structure interaction (FSI) simulations.

physics.flu-dyn↗

arXiv2025

CausalSR: Structural Causal Model-Driven Super-Resolution with Counterfactual Inference

Physical and optical factors interacting with sensor characteristics create complex image degradation patterns. Despite advances in deep learning-based super-resolution, existing methods overlook the causal nature of degradation by adopting simplistic black-box mappings. This paper formulates super-resolution using structural causal models to reason about image degradation processes. We establish a mathematical foundation that unifies principles from causal inference, deriving necessary conditions for identifying latent degradation mechanisms and corresponding propagation. We propose a novel counterfactual learning strategy that leverages semantic guidance to reason about hypothetical degradation scenarios, leading to theoretically-grounded representations that capture invariant features across different degradation conditions. The framework incorporates an adaptive intervention mechanism with provable bounds on treatment effects, allowing precise manipulation of degradation factors while maintaining semantic consistency. Through extensive empirical validation, we demonstrate that our approach achieves significant improvements over state-of-the-art methods, particularly in challenging scenarios with compound degradations. On standard benchmarks, our method consistently outperforms existing approaches by significant margins (0.86-1.21dB PSNR), while providing interpretable insights into the restoration process. The theoretical framework and empirical results demonstrate the fundamental importance of causal reasoning in understanding image restoration systems.

cs.CV↗

arXiv2025

"Molecular waveplate" for the control of ultrashort pulses carrying orbital angular momentum

Ultrashort laser pulses carrying orbital angular momentum (OAM) have become essential tools in Atomic, Molecular, and Optical (AMO) studies, particularly for investigating strong-field light-matter interactions. However, controlling and generating ultrashort vortex pulses presents significant challenges, since their broad spectral content complicates manipulation with conventional optical elements, while the high peak power inherent in short-duration pulses risks damaging optical components. Here, we introduce a novel method for generating and controlling broadband ultrashort vortex beams by exploiting the non-adiabatic alignment of linear gas-phase molecules induced by vector beams. The interaction between the vector beam and the gas-phase molecules results in spatially varying polarizability, imparting a phase modulation to a probe laser. This process effectively creates a tunable ``molecular waveplate'' that adapts naturally to a broad spectral range. By leveraging this approach, we can generate ultrashort vortex pulses across a wide range of wavelengths. Under optimized gas pressure and interaction length conditions, this method allows for highly efficient conversion of circularly polarized light into the desired OAM pulse, thus enabling the generation of few-cycle, high-intensity vortex beams. This molecular waveplate, which overcomes the limitations imposed by conventional optical elements, opens up new possibilities for exploring strong-field physics, ultrafast science, and other applications that require high-intensity vortex beams.

physics.optics↗