arXiv subjects

Feng Wu

Publications and source records attributed to Feng Wu.

At least 91 records · Page 5Linked to original sources

arXiv2025

Label Deconvolution for Node Representation Learning on Large-scale Attributed Graphs against Learning Bias

Node representation learning on attributed graphs -- whose nodes are associated with rich attributes (e.g., texts and protein sequences) -- plays a crucial role in many important downstream tasks. To encode the attributes and graph structures simultaneously, recent studies integrate pre-trained models with graph neural networks (GNNs), where pre-trained models serve as node encoders (NEs) to encode the attributes. As jointly training large NEs and GNNs on large-scale graphs suffers from severe scalability issues, many methods propose to train NEs and GNNs separately. Consequently, they do not take feature convolutions in GNNs into consideration in the training phase of NEs, leading to a significant learning bias relative to the joint training. To address this challenge, we propose an efficient label regularization technique, namely Label Deconvolution (LD), to alleviate the learning bias by a novel and highly scalable approximation to the inverse mapping of GNNs. The inverse mapping leads to an objective function that is equivalent to that by the joint training, while it can effectively incorporate GNNs in the training phase of NEs against the learning bias. More importantly, we show that LD converges to the optimal objective function values by the joint training under mild assumptions. Experiments demonstrate LD significantly outperforms state-of-the-art methods on Open Graph Benchmark datasets.

cs.LG↗

arXiv2025

HATA: Trainable and Hardware-Efficient Hash-Aware Top-k Attention for Scalable Large Model Inference

Large Language Models (LLMs) have emerged as a pivotal research area, yet the attention module remains a critical bottleneck in LLM inference, even with techniques like KVCache to mitigate redundant computations. While various top-$k$ attention mechanisms have been proposed to accelerate LLM inference by exploiting the inherent sparsity of attention, they often struggled to strike a balance between efficiency and accuracy. In this paper, we introduce HATA (Hash-Aware Top-$k$ Attention), a novel approach that systematically integrates low-overhead learning-to-hash techniques into the Top-$k$ attention process. Different from the existing top-k attention methods which are devoted to seeking an absolute estimation of qk score, typically with a great cost, HATA maps queries and keys into binary hash codes, and acquires the relative qk score order with a quite low cost, which is sufficient for realizing top-k attention. Extensive experiments demonstrate that HATA achieves up to 7.2$\times$ speedup compared to vanilla full attention while maintaining model accuracy. In addition, HATA outperforms the state-of-the-art top-$k$ attention methods in both accuracy and efficiency across multiple mainstream LLM models and diverse tasks. HATA is open source at https://github.com/gpzlx1/HATA.

cs.LG↗

arXiv2025

HyperTree Planning: Enhancing LLM Reasoning via Hierarchical Thinking

Recent advancements have significantly enhanced the performance of large language models (LLMs) in tackling complex reasoning tasks, achieving notable success in domains like mathematical and logical reasoning. However, these methods encounter challenges with complex planning tasks, primarily due to extended reasoning steps, diverse constraints, and the challenge of handling multiple distinct sub-tasks. To address these challenges, we propose HyperTree Planning (HTP), a novel reasoning paradigm that constructs hypertree-structured planning outlines for effective planning. The hypertree structure enables LLMs to engage in hierarchical thinking by flexibly employing the divide-and-conquer strategy, effectively breaking down intricate reasoning steps, accommodating diverse constraints, and managing multiple distinct sub-tasks in a well-organized manner. We further introduce an autonomous planning framework that completes the planning process by iteratively refining and expanding the hypertree-structured planning outlines. Experiments demonstrate the effectiveness of HTP, achieving state-of-the-art accuracy on the TravelPlanner benchmark with Gemini-1.5-Pro, resulting in a 3.6 times performance improvement over o1-preview.

cs.AI↗

arXiv2025

Accelerating Large Language Model Reasoning via Speculative Search

Tree-search-based reasoning methods have significantly enhanced the reasoning capability of large language models (LLMs) by facilitating the exploration of multiple intermediate reasoning steps, i.e., thoughts. However, these methods suffer from substantial inference latency, as they have to generate numerous reasoning thoughts, severely limiting LLM applicability. To address this challenge, we propose a novel Speculative Search (SpecSearch) framework that significantly accelerates LLM reasoning by optimizing thought generation. Specifically, SpecSearch utilizes a small model to strategically collaborate with a large model at both thought and token levels, efficiently generating high-quality reasoning thoughts. The major pillar of SpecSearch is a novel quality-preserving rejection mechanism, which effectively filters out thoughts whose quality falls below that of the large model's outputs. Moreover, we show that SpecSearch preserves comparable reasoning quality to the large model. Experiments on both the Qwen and Llama models demonstrate that SpecSearch significantly outperforms state-of-the-art approaches, achieving up to 2.12$\times$ speedup with comparable reasoning quality.

cs.CL↗

arXiv2025

Effects of off-diagonal permittivity terms on polarization singularities in anisotropic grating system

The evolutions of polarization singularities, including bound states in the continuum (BICs) and circularly polarized states (C points), are usually realized by tuning the geometric parameters of photonic crystal slabs. Here, we use the off-diagonal terms of permittivity tensor to manipulate polarization singularities without breaking the structural symmetry in an anisotropic grating system. By controlling the optical axis of anisotropic media, BICs can be shifted to different positions or split into C points, meanwhile, the creation and annihilation of multiple C points are also observed during the evolution process for both TE and TM modes, respectively. Remarkably, two different splitting directions of BICs can be achieved by tuning the off-diagonal terms of permittivity tensor for the two modes. This work illustrates the important role of off-diagonal terms on the far-field polarization singularities and provide an alternative way to precisely manipulate optical singularities

physics.optics↗

arXiv2025

Seismic analysis based on a new interval method with incomplete information

For seismic analysis in engineering structures, it is essential to consider the dynamic responses under seismic excitation, necessitating the description of seismic accelerations. Limit seismics samples lead to incomplete uncertainty information, which is described by the non-probabilistic method reasonable. This study employs the minimum interval radius-based interval process (MRIP) based on the convex model to describe the time-variant uncertain seismic acceleration, subsequently conducting uncertainty analysis for seismic structures. However, the Monte Carlo simulation for uncertainty analysis requires extensive deterministic computations to ensure accuracy, exhibiting poor computational efficiency. To address this issue, this paper first improves the covariance matrix adaptation evolution strategy (CMA-ES) through the dynamic evolution sequence, proposing DES-ES, whose efficiency is validated to be higher than that of CMA-ES. Furthermore, leveraging the dependency of the responses, a computational framework named DES-ES-SS is proposed. Numerical experiments demonstrate that DES-ES-SS improves computational efficiency while maintaining the accuracy of the interval uncertainty analysis of the seismic structures whether the seismic acceleration is stationary or non-stationary.

cs.CE↗

arXiv2025

High coherence fluxonium manufactured with a wafer-scale uniformity process

Fluxonium qubits are recognized for their high coherence times and high operation fidelities, attributed to their unique design incorporating a superinductor, which is typically implemented using an array of over 100 Josephson junctions. However, this complexity poses significant fabrication challenges, particularly in achieving high yield and junction uniformity with traditional methods. Here, we introduce an overlap process for Josephson junction fabrication that achieves nearly 100% yield and maintains uniformity across a 2-inch wafer with less than 5% variation for the phase slip junction and less than 2% for the entire junction array. We use a compact junction array design that achieves state-of-the-art dielectric loss tangents and flux noise levels, as confirmed by multiple devices. This enables fluxonium qubits to reach energy relaxation times exceeding 1 millisecond at the flux frustration point. This work paves the way for scalable high coherence fluxonium processors using CMOS-compatible processes, marking a significant step towards practical quantum computing.

quant-ph↗

arXiv2025

Electro-optically tunable second-harmonic generation in lithium niobate metasurfaces boosted by Brillouin zone folding induced bound states in the continuum

Lithium niobate (LN) metasurfaces exhibit remarkable Pockels effect-driven electro-optic tunability, enabling dynamic control of optical responses through external electric fields. When combined with their high second-order nonlinear susceptibility ($χ^{2}$), this tunability is projected into the nonlinear landscape, realizing second-harmonic generation (SHG)-dominated functionalities in integrated photonics. However, achieving deep SHG modulation in LN metasurfaces remains challenging due to LN's limited refractive index tunability under practical driving voltage. To address this, we design an air-hole-structured LN metasurface by strategically adjusting air hole positions to induce Brillouin zone folding-enabled bound states in the continuum with ultrahigh quality factors ($Q>10^{4}$). Numerical simulations demonstrate a $2.59\%$ SHG conversion efficiency at 0.1 MW/cm$^{2}$ excitation and a modulation depth exceeding 0.99 under 15 V peak-to-peak voltage ($ΔV_{\text{pp}}$). This work establishes a compact framework for electrically tunable nonlinear optics, advancing applications in integrated quantum light sources and programmable photonic chips.

physics.optics↗

arXiv2025

SuperGrad: a differentiable simulator for superconducting processors

One significant advantage of superconducting processors is their extensive design flexibility, which encompasses various types of qubits and interactions. Given the large number of tunable parameters of a processor, the ability to perform gradient optimization would be highly beneficial. Efficient backpropagation for gradient computation requires a tightly integrated software library, for which no open-source implementation is currently available. In this work, we introduce SuperGrad, a simulator that accelerates the design of superconducting quantum processors by incorporating gradient computation capabilities. SuperGrad offers a user-friendly interface for constructing Hamiltonians and computing both static and dynamic properties of composite systems. This differentiable simulation is valuable for a range of applications, including optimal control, design optimization, and experimental data fitting. In this paper, we demonstrate these applications through examples and code snippets.

quant-ph↗

arXiv2025

The effects of disorder in superconducting materials on qubit coherence

Introducing disorderness in the superconducting materials has been considered promising to enhance the electromagnetic impedance and realize noise-resilient superconducting qubits. Despite a number of pioneering implementations, the understanding of the correlation between the material disorderness and the qubit coherence is still developing. Here, we demonstrate a systematic characterization of fluxonium qubits with the superinductors made from titanium-aluminum-nitride with varied disorderness. From qubit noise spectroscopy, the flux noise and the dielectric loss are extracted as a measure of the coherence properties. Our results reveal that the $1/f$ flux noise dominates the qubit decoherence around the flux-frustration point, strongly correlated with the material disorderness; while the dielectric loss remains low under a wide range of material properties. From the flux-noise amplitudes, the areal density ($σ$) of the phenomenological spin defects and material disorderness are found to be approximately correlated by $σ\propto ρ_{xx}^3$, or effectively $(k_F l)^{-3}$. This work has provided new insights on the origin of decoherence channels within superconductors, and could serve as a useful guideline for material design and optimization.

quant-ph↗

arXiv2025

CLCR: Contrastive Learning-based Constraint Reordering for Efficient MILP Solving

Constraint ordering plays a critical role in the efficiency of Mixed-Integer Linear Programming (MILP) solvers, particularly for large-scale problems where poorly ordered constraints trigger increased LP iterations and suboptimal search trajectories. This paper introduces CLCR (Contrastive Learning-based Constraint Reordering), a novel framework that systematically optimizes constraint ordering to accelerate MILP solving. CLCR first clusters constraints based on their structural patterns and then employs contrastive learning with a pointer network to optimize their sequence, preserving problem equivalence while improving solver efficiency. Experiments on benchmarks show CLCR reduces solving time by 30% and LP iterations by 25% on average, without sacrificing solution accuracy. This work demonstrates the potential of data-driven constraint ordering to enhance optimization models, offering a new paradigm for bridging mathematical programming with machine learning.

cs.LG↗

arXiv2025

Beyond Local Selection: Global Cut Selection for Enhanced Mixed-Integer Programming

In mixed-integer programming (MIP) solvers, cutting planes are essential for Branch-and-Cut (B&C) algorithms as they reduce the search space and accelerate the solving process. Traditional methods rely on hard-coded heuristics for cut plane selection but fail to leverage problem-specific structural features. Recent machine learning approaches use neural networks for cut selection but focus narrowly on the efficiency of single-node within the B&C algorithm, without considering the broader contextual information. To address this, we propose Global Cut Selection (GCS), which uses a bipartite graph to represent the search tree and combines graph neural networks with reinforcement learning to develop cut selection strategies. Unlike prior methods, GCS applies cutting planes across all nodes, incorporating richer contextual information. Experiments show GCS significantly improves solving efficiency for synthetic and large-scale real-world MIPs compared to traditional and learning-based methods.

cs.AI↗

arXiv2025

USTC-TD: A Test Dataset and Benchmark for Image and Video Coding in 2020s

Image/video coding has been a remarkable research area for both academia and industry for many years. Testing datasets, especially high-quality image/video datasets are desirable for the justified evaluation of coding-related research, practical applications, and standardization activities. We put forward a test dataset namely USTC-TD, which has been successfully adopted in the practical end-to-end image/video coding challenge of the IEEE International Conference on Visual Communications and Image Processing (VCIP) in 2022 and 2023. USTC-TD contains 40 images at 4K spatial resolution and 10 video sequences at 1080p spatial resolution, featuring various content due to the diverse environmental factors (e.g. scene type, texture, motion, view) and the designed imaging factors (e.g. illumination, lens, shadow). We quantitatively evaluate USTC-TD on different image/video features (spatial, temporal, color, lightness), and compare it with the previous image/video test datasets, which verifies its excellent compensation for the shortcomings of existing datasets. We also evaluate both classic standardized and recently learned image/video coding schemes on USTC-TD using objective quality metrics (PSNR, MS-SSIM, VMAF) and subjective quality metric (MOS), providing an extensive benchmark for these evaluated schemes. Based on the characteristics and specific design of the proposed test dataset, we analyze the benchmark performance and shed light on the future research and development of image/video coding. All the data are released online: https://esakak.github.io/USTC-TD.

eess.IV↗

arXiv2025

Structural and Statistical Texture Knowledge Distillation and Learning for Segmentation

Low-level texture feature/knowledge is also of vital importance for characterizing the local structural pattern and global statistical properties, such as boundary, smoothness, regularity, and color contrast, which may not be well addressed by high-level deep features. In this paper, we aim to re-emphasize the low-level texture information in deep networks for semantic segmentation and related knowledge distillation tasks. To this end, we take full advantage of both structural and statistical texture knowledge and propose a novel Structural and Statistical Texture Knowledge Distillation (SSTKD) framework for semantic segmentation. Specifically, Contourlet Decomposition Module (CDM) is introduced to decompose the low-level features with iterative Laplacian pyramid and directional filter bank to mine the structural texture knowledge, and Texture Intensity Equalization Module (TIEM) is designed to extract and enhance the statistical texture knowledge with the corresponding Quantization Congruence Loss (QDL). Moreover, we propose the Co-occurrence TIEM (C-TIEM) and generic segmentation frameworks, namely STLNet++ and U-SSNet, to enable existing segmentation networks to harvest the structural and statistical texture information more effectively. Extensive experimental results on three segmentation tasks demonstrate the effectiveness of the proposed methods and their state-of-the-art performance on seven popular benchmark datasets, respectively.

cs.CV↗

arXiv2025

Apollo-MILP: An Alternating Prediction-Correction Neural Solving Framework for Mixed-Integer Linear Programming

Leveraging machine learning (ML) to predict an initial solution for mixed-integer linear programming (MILP) has gained considerable popularity in recent years. These methods predict a solution and fix a subset of variables to reduce the problem dimension. Then, they solve the reduced problem to obtain the final solutions. However, directly fixing variable values can lead to low-quality solutions or even infeasible reduced problems if the predicted solution is not accurate enough. To address this challenge, we propose an Alternating prediction-correction neural solving framework (Apollo-MILP) that can identify and select accurate and reliable predicted values to fix. In each iteration, Apollo-MILP conducts a prediction step for the unfixed variables, followed by a correction step to obtain an improved solution (called reference solution) through a trust-region search. By incorporating the predicted and reference solutions, we introduce a novel Uncertainty-based Error upper BOund (UEBO) to evaluate the uncertainty of the predicted values and fix those with high confidence. A notable feature of Apollo-MILP is the superior ability for problem reduction while preserving optimality, leading to high-quality final solutions. Experiments on commonly used benchmarks demonstrate that our proposed Apollo-MILP significantly outperforms other ML-based approaches in terms of solution quality, achieving over a 50% reduction in the solution gap.

cs.LG↗

arXiv2025

PyramidDrop: Accelerating Your Large Vision-Language Models via Pyramid Visual Redundancy Reduction

In large vision-language models (LVLMs), images serve as inputs that carry a wealth of information. As the idiom "A picture is worth a thousand words" implies, representing a single image in current LVLMs can require hundreds or even thousands of tokens. This results in significant computational costs, which grow quadratically as input image resolution increases, thereby severely impacting the efficiency of both training and inference. Previous approaches have attempted to reduce the number of image tokens either before or within the early layers of LVLMs. However, these strategies inevitably result in the loss of crucial image information, ultimately diminishing model performance. To address this challenge, we conduct an empirical study revealing that all visual tokens are necessary for LVLMs in the shallow layers, and token redundancy progressively increases in the deeper layers of the model. To this end, we propose PyramidDrop, a visual redundancy reduction strategy for LVLMs to boost their efficiency in both training and inference with neglectable performance loss. Specifically, we partition the LVLM into several stages and drop part of the image tokens at the end of each stage with a pre-defined ratio, creating pyramid-like visual tokens across model layers. The dropping is based on a lightweight similarity calculation with a negligible time overhead. Extensive experiments demonstrate that PyramidDrop can achieve a 40% training time and 55% inference FLOPs acceleration of LLaVA-NeXT with comparable performance. Besides, the PyramidDrop could also serve as a plug-and-play strategy for inference acceleration without training, with better performance and lower inference cost than counterparts. Code is available at https://github.com/Cooperx521/PyramidDrop.

cs.CV↗

arXiv2025

Accurate and Scalable Graph Neural Networks via Message Invariance

Message passing-based graph neural networks (GNNs) have achieved great success in many real-world applications. For a sampled mini-batch of target nodes, the message passing process is divided into two parts: message passing between nodes within the batch (MP-IB) and message passing from nodes outside the batch to those within it (MP-OB). However, MP-OB recursively relies on higher-order out-of-batch neighbors, leading to an exponentially growing computational cost with respect to the number of layers. Due to the neighbor explosion, the whole message passing stores most nodes and edges on the GPU such that many GNNs are infeasible to large-scale graphs. To address this challenge, we propose an accurate and fast mini-batch approach for large graph transductive learning, namely topological compensation (TOP), which obtains the outputs of the whole message passing solely through MP-IB, without the costly MP-OB. The major pillar of TOP is a novel concept of message invariance, which defines message-invariant transformations to convert costly MP-OB into fast MP-IB. This ensures that the modified MP-IB has the same output as the whole message passing. Experiments demonstrate that TOP is significantly faster than existing mini-batch methods by order of magnitude on vast graphs (millions of nodes and billions of edges) with limited accuracy degradation.

cs.LG↗

arXiv2025

Extracting intrinsic superconducting properties in intercalated layered superconductors using an extended 2D Tinkham model

Bulk two dimensional (2D) superconductivity has gained considerable attention due to its intricate interplay between symmetry breaking, nontrivial topology, 2D phase fluctuations, and unconventional superconductivity. However, certain intercalated layered superconductors, despite their short c-axis superconducting coherence length, have been misclassified as anisotropic three-dimensional (3D) superconductors. Here, we investigate (Li,Fe)OHFeSe superconductors with varying degrees of interlayer misalignment, revealing sample-dependent superconducting dimensionality while consistently observing Berezinskii Kosterlitz Thouless (BKT) transitions. To resolve this discrepancy, we develop an extended 2D Tinkham model that quantitatively captures the blurring effects induced by interlayer misalignment. We further demonstrate the validity of this model in both (Li,Fe)OHFeSe and cetyltrimethyl ammonium (CTA+) intercalated (CTA)0.5SnSe2 superconductors, highlighting its broad applicability. This work provides valuable insights into bulk 2D superconductivity and establishes an extended 2D Tinkham model for quantitatively extracting intrinsic superconducting properties in intercalated layered superconductors, particularly those exhibiting significant interlayer misalignments.

cond-mat.supr-con↗