arXiv subjects

Yu Wang

Publications and source records attributed to Yu Wang.

At least 505 records · Page 28Linked to original sources

arXiv2025

ZPD-SCA: Unveiling the Blind Spots of LLMs in Assessing Students' Cognitive Abilities

Large language models (LLMs) have demonstrated potential in educational applications, yet their capacity to accurately assess the cognitive alignment of reading materials with students' developmental stages remains insufficiently explored. This gap is particularly critical given the foundational educational principle of the Zone of Proximal Development (ZPD), which emphasizes the need to match learning resources with Students' Cognitive Abilities (SCA). Despite the importance of this alignment, there is a notable absence of comprehensive studies investigating LLMs' ability to evaluate reading comprehension difficulty across different student age groups, especially in the context of Chinese language education. To fill this gap, we introduce ZPD-SCA, a novel benchmark specifically designed to assess stage-level Chinese reading comprehension difficulty. The benchmark is annotated by 60 Special Grade teachers, a group that represents the top 0.15% of all in-service teachers nationwide. Experimental results reveal that LLMs perform poorly in zero-shot learning scenarios, with Qwen-max and GLM even falling below the probability of random guessing. When provided with in-context examples, LLMs performance improves substantially, with some models achieving nearly double the accuracy of their zero-shot baselines. These results reveal that LLMs possess emerging abilities to assess reading difficulty, while also exposing limitations in their current training for educationally aligned judgment. Notably, even the best-performing models display systematic directional biases, suggesting difficulties in accurately aligning material difficulty with SCA. Furthermore, significant variations in model performance across different genres underscore the complexity of task. We envision that ZPD-SCA can provide a foundation for evaluating and improving LLMs in cognitively aligned educational applications.

cs.CL↗

arXiv2025

Data-Driven Analysis and Predictive Control of Descriptor Systems with Application to Power and Water Networks

Despite growing interest in data-driven analysis and control of linear systems, descriptor systems--which are essential for modeling complex engineered systems with algebraic constraints like power and water networks--have received comparatively little attention. This paper develops a comprehensive data-driven framework for analyzing and controlling discrete-time descriptor systems without relying on explicit state-space models. We address fundamental challenges posed by non-causality through the construction of forward and backward data matrices, establishing data-based sufficient conditions for controllability and observability in terms of input-output data, where both R-controllability and C-controllability (R-observability and C-observability) have been considered. We then extend Willems' fundamental lemma to incompletely controllable systems. These methodological advances enable Data-Enabled Predictive Control (DeePC) to achieve output tracking in descriptor systems and to maintain performance under incomplete controllability conditions, as demonstrated in two case studies: i) Frequency regulation in an IEEE 9-bus power system with 3 generators, where DeePC maintained the frequency stability of the power system despite deliberate violations of R-controllability; and ii) Pressure head control in an EPANET water network with 3 tanks, 2 reservoirs, and 117 pipes, where output tracking was successfully enforced under algebraic constraints.

eess.SY↗

arXiv2025

Task-Generalized Adaptive Cross-Domain Learning for Multimodal Image Fusion

Multimodal Image Fusion (MMIF) aims to integrate complementary information from different imaging modalities to overcome the limitations of individual sensors. It enhances image quality and facilitates downstream applications such as remote sensing, medical diagnostics, and robotics. Despite significant advancements, current MMIF methods still face challenges such as modality misalignment, high-frequency detail destruction, and task-specific limitations. To address these challenges, we propose AdaSFFuse, a novel framework for task-generalized MMIF through adaptive cross-domain co-fusion learning. AdaSFFuse introduces two key innovations: the Adaptive Approximate Wavelet Transform (AdaWAT) for frequency decoupling, and the Spatial-Frequency Mamba Blocks for efficient multimodal fusion. AdaWAT adaptively separates the high- and low-frequency components of multimodal images from different scenes, enabling fine-grained extraction and alignment of distinct frequency characteristics for each modality. The Spatial-Frequency Mamba Blocks facilitate cross-domain fusion in both spatial and frequency domains, enhancing this process. These blocks dynamically adjust through learnable mappings to ensure robust fusion across diverse modalities. By combining these components, AdaSFFuse improves the alignment and integration of multimodal features, reduces frequency loss, and preserves critical details. Extensive experiments on four MMIF tasks -- Infrared-Visible Image Fusion (IVF), Multi-Focus Image Fusion (MFF), Multi-Exposure Image Fusion (MEF), and Medical Image Fusion (MIF) -- demonstrate AdaSFFuse's superior fusion performance, ensuring both low computational cost and a compact network, offering a strong balance between performance and efficiency. The code will be publicly available at https://github.com/Zhen-yu-Liu/AdaSFFuse.

cs.CV↗

arXiv2025

Img2ST-Net: Efficient High-Resolution Spatial Omics Prediction from Whole Slide Histology Images via Fully Convolutional Image-to-Image Learning

Recent advances in multi-modal AI have demonstrated promising potential for generating the currently expensive spatial transcriptomics (ST) data directly from routine histology images, offering a means to reduce the high cost and time-intensive nature of ST data acquisition. However, the increasing resolution of ST, particularly with platforms such as Visium HD achieving 8um or finer, introduces significant computational and modeling challenges. Conventional spot-by-spot sequential regression frameworks become inefficient and unstable at this scale, while the inherent extreme sparsity and low expression levels of high-resolution ST further complicate both prediction and evaluation. To address these limitations, we propose Img2ST-Net, a novel histology-to-ST generation framework for efficient and parallel high-resolution ST prediction. Unlike conventional spot-by-spot inference methods, Img2ST-Net employs a fully convolutional architecture to generate dense, HD gene expression maps in a parallelized manner. By modeling HD ST data as super-pixel representations, the task is reformulated from image-to-omics inference into a super-content image generation problem with hundreds or thousands of output channels. This design not only improves computational efficiency but also better preserves the spatial organization intrinsic to spatial omics data. To enhance robustness under sparse expression patterns, we further introduce SSIM-ST, a structural-similarity-based evaluation metric tailored for high-resolution ST analysis. We present a scalable, biologically coherent framework for high-resolution ST prediction. Img2ST-Net offers a principled solution for efficient and accurate ST inference at scale. Our contributions lay the groundwork for next-generation ST modeling that is robust and resolution-aware. The source code has been made publicly available at https://github.com/hrlblab/Img2ST-Net.

cs.CV↗

arXiv2025

Shadows of three black holes in static equilibrium configuration

In this paper, we employ a ray-tracing algorithm to simulate the shadows of three equal-mass black holes in static equilibrium across a wide parameter space. We find that the shadows consist of a larger primary shadow and several distorted, eyebrow-like secondary shadows. The boundaries of these profiles exhibit self-similar fractal structures, which can be attributed to the photon chaotic scattering. In certain parameter spaces, we also observe the ring-like shadows, with the ring diameter associated with the spacing of black holes. Furthermore, when the black holes approach each other sufficiently, their shadows can merge into a standard disk, suggesting a shadow degeneracy between closely arranged triple black holes and a single massive, spherically symmetric black hole. The shadow features of the triple black holes revealed in this study have potential implications for analyzing the shadow formation mechanisms, as well as the gravitational lensing during the merger and inspiral of black holes.

gr-qc↗

arXiv2025

Enhanced Krylov Methods for Molecular Hamiltonians: Reduced Memory Cost and Complexity Scaling via Tensor Hypercontraction

We introduce an algorithm that is simultaneously memory-efficient and low-scaling for applying ab initio molecular Hamiltonians to matrix-product states (MPS) via the tensor-hypercontraction (THC) format. These gains carry over to Krylov subspace methods, which can find low-lying eigenstates and simulate quantum time evolution while avoiding local minima and maintaining high accuracy. In our approach, the molecular Hamiltonian is represented as a sum of products of four MPOs, each with a bond dimension of only 2. Iteratively applying the MPOs to the current quantum state in MPS form, summing and re-compressing the MPS leads to a scheme with the same asymptotic memory cost as the bare MPS and reduces the computational cost scaling compared to the Krylov method using a conventional MPO construction. We provide a detailed theoretical derivation of these statements and conduct supporting numerical experiments to demonstrate the advantage. Our algorithm is highly parallelizable and thus lends itself to large-scale HPC simulations.

cond-mat.str-el↗

arXiv2025

Multi-Sequence Parotid Gland Lesion Segmentation via Expert Text-Guided Segment Anything Model

Parotid gland lesion segmentation is essential for the treatment of parotid gland diseases. However, due to the variable size and complex lesion boundaries, accurate parotid gland lesion segmentation remains challenging. Recently, the Segment Anything Model (SAM) fine-tuning has shown remarkable performance in the field of medical image segmentation. Nevertheless, SAM's interaction segmentation model relies heavily on precise lesion prompts (points, boxes, masks, etc.), which are very difficult to obtain in real-world applications. Besides, current medical image segmentation methods are automatically generated, ignoring the domain knowledge of medical experts when performing segmentation. To address these limitations, we propose the parotid gland segment anything model (PG-SAM), an expert diagnosis text-guided SAM incorporating expert domain knowledge for cross-sequence parotid gland lesion segmentation. Specifically, we first propose an expert diagnosis report guided prompt generation module that can automatically generate prompt information containing the prior domain knowledge to guide the subsequent lesion segmentation process. Then, we introduce a cross-sequence attention module, which integrates the complementary information of different modalities to enhance the segmentation effect. Finally, the multi-sequence image features and generated prompts are feed into the decoder to get segmentation result. Experimental results demonstrate that PG-SAM achieves state-of-the-art performance in parotid gland lesion segmentation across three independent clinical centers, validating its clinical applicability and the effectiveness of diagnostic text for enhancing image segmentation in real-world clinical settings.

cs.CV↗

arXiv2025

D3P: Dynamic Denoising Diffusion Policy via Reinforcement Learning

Diffusion policies excel at learning complex action distributions for robotic visuomotor tasks, yet their iterative denoising process poses a major bottleneck for real-time deployment. Existing acceleration methods apply a fixed number of denoising steps per action, implicitly treating all actions as equally important. However, our experiments reveal that robotic tasks often contain a mix of \emph{crucial} and \emph{routine} actions, which differ in their impact on task success. Motivated by this finding, we propose \textbf{D}ynamic \textbf{D}enoising \textbf{D}iffusion \textbf{P}olicy \textbf{(D3P)}, a diffusion-based policy that adaptively allocates denoising steps across actions at test time. D3P uses a lightweight, state-aware adaptor to allocate the optimal number of denoising steps for each action. We jointly optimize the adaptor and base diffusion policy via reinforcement learning to balance task performance and inference efficiency. On simulated tasks, D3P achieves an averaged 2.2$\times$ inference speed-up over baselines without degrading success. Furthermore, we demonstrate D3P's effectiveness on a physical robot, achieving a 1.9$\times$ acceleration over the baseline.

cs.RO↗

arXiv2025

CityLight: A Neighborhood-inclusive Universal Model for Coordinated City-scale Traffic Signal Control

City-scale traffic signal control (TSC) involves thousands of heterogeneous intersections with varying topologies, making cooperative decision-making across intersections particularly challenging. Given the prohibitive computational cost of learning individual policies for each intersection, some researchers explore learning a universal policy to control each intersection in a decentralized manner, where the key challenge is to construct a universal representation method for heterogeneous intersections. However, existing methods are limited to universally representing information of heterogeneous ego intersections, neglecting the essential representation of influence from their heterogeneous neighbors. Universally incorporating neighborhood information is nontrivial due to the intrinsic complexity of traffic flow interactions, as well as the challenge of modeling collective influences from neighbor intersections. To address these challenges, we propose CityLight, which learns a universal policy based on representations obtained with two major modules: a Neighbor Influence Encoder to explicitly model neighbor's influence with specified traffic flow relation and connectivity to the ego intersection; a Neighbor Influence Aggregator to attentively aggregate the influence of neighbors based on their mutual competitive relations. Extensive experiments on five city-scale datasets, ranging from 97 to 13,952 intersections, confirm the efficacy of CityLight, with an average throughput improvement of 11.68% and a lift of 22.59% for generalization.

eess.SY↗

arXiv2025

Multiscale Coupled Polarization and BKT Transitions in Tow-Dimensional Hybrid Organic-Inorganic Perovskites

We present an extended two-dimensional XY rotor model specifically designed to capture the polarization dynamics of hybrid organic-inorganic perovskite monolayers. This framework integrates nearest and next-nearest neighbor couplings, crystalline anisotropy inherent to perovskite lattice symmetries, external bias fields, and long-range dipolar interactions that are prominent in layered perovskite architectures. Through a combination of analytical coarse-graining and large-scale Monte Carlo simulations on 64*64 lattices, we identify two distinct thermodynamic regimes: a low-temperature quasi-ferroelectric state characterized by finite polarization and domain wall formation, and a higher-temperature Berezinskii-Kosterlitz-Thouless (BKT) crossover associated with vortex-antivortex unbinding and the suppression of long-range order. Our results reproduce key experimental signatures observed in quasi-two-dimensional perovskites, including dual peaks in dielectric susceptibility, enhanced vortex density near the transition, multistable polarization hysteresis under applied fields, and the scaling behavior of domain wall widths. This minimal yet realistic model provides a unifying perspective on how topological transitions and ferroelectric ordering coexist in layered perovskite systems, offering quantitative guidance for interpreting the emergent polar vortex lattices and complex phase behavior recently reported in hybrid perovskite thin films.

cond-mat.mtrl-sci↗

arXiv2025

PLGS: Robust Panoptic Lifting with 3D Gaussian Splatting

Previous methods utilize the Neural Radiance Field (NeRF) for panoptic lifting, while their training and rendering speed are unsatisfactory. In contrast, 3D Gaussian Splatting (3DGS) has emerged as a prominent technique due to its rapid training and rendering speed. However, unlike NeRF, the conventional 3DGS may not satisfy the basic smoothness assumption as it does not rely on any parameterized structures to render (e.g., MLPs). Consequently, the conventional 3DGS is, in nature, more susceptible to noisy 2D mask supervision. In this paper, we propose a new method called PLGS that enables 3DGS to generate consistent panoptic segmentation masks from noisy 2D segmentation masks while maintaining superior efficiency compared to NeRF-based methods. Specifically, we build a panoptic-aware structured 3D Gaussian model to introduce smoothness and design effective noise reduction strategies. For the semantic field, instead of initialization with structure from motion, we construct reliable semantic anchor points to initialize the 3D Gaussians. We then use these anchor points as smooth regularization during training. Additionally, we present a self-training approach using pseudo labels generated by merging the rendered masks with the noisy masks to enhance the robustness of PLGS. For the instance field, we project the 2D instance masks into 3D space and match them with oriented bounding boxes to generate cross-view consistent instance masks for supervision. Experiments on various benchmarks demonstrate that our method outperforms previous state-of-the-art methods in terms of both segmentation quality and speed.

cs.CV↗

arXiv2025

Doubling property of self-similar measures with overlaps

Recently, Yang, Yuan and Zhang [Doubling properties of self-similar measures and Bernoulli measures on self-affine Sierpinski sponges, Indiana Univ. Math. J., 73 (2024), 475-492] characterized when a self-similar measure satisfying the open set condition is doubling. In this paper, we study when a self-similar measure with overlaps is doubling. Let $m\geq 2$ and let $β>1$ be the Pisot number satisfying $β^m=\sum_{j=0}^{m-1}β^j$. Let $\mathbf{p}=(p_1,p_2)$ be a probability weight and let $μ_{\mathbf{p}}$ be the self-similar measure associated to the IFS $\{ S_1(x)={x}/β, S_2(x)={x}/β+(1-{1}/β),\}.$ Yung [...,Indiana Univ. Math. J., ] proved that when $m=2$, $μ_{\mathbf{p}}$ is doubling if and only if $\mathbf{p}=(1/2,1/2)$. We show that for $m\geq 3$, $μ_{\mathbf{p}}$ is always non-doubling.

math.MG↗

arXiv2025

DHCP: Detecting Hallucinations by Cross-modal Attention Pattern in Large Vision-Language Models

Large vision-language models (LVLMs) have demonstrated exceptional performance on complex multimodal tasks. However, they continue to suffer from significant hallucination issues, including object, attribute, and relational hallucinations. To accurately detect these hallucinations, we investigated the variations in cross-modal attention patterns between hallucination and non-hallucination states. Leveraging these distinctions, we developed a lightweight detector capable of identifying hallucinations. Our proposed method, Detecting Hallucinations by Cross-modal Attention Patterns (DHCP), is straightforward and does not require additional LVLM training or extra LVLM inference steps. Experimental results show that DHCP achieves remarkable performance in hallucination detection. By offering novel insights into the identification and analysis of hallucinations in LVLMs, DHCP contributes to advancing the reliability and trustworthiness of these models. The code is available at https://github.com/btzyd/DHCP.

cs.CV↗

arXiv2025

DEPTHOR: Depth Enhancement from a Practical Light-Weight dToF Sensor and RGB Image

Depth enhancement, which uses RGB images as guidance to convert raw signals from dToF into high-precision, dense depth maps, is a critical task in computer vision. Although existing super-resolution-based methods show promising results on public datasets, they often rely on idealized assumptions like accurate region correspondences and reliable dToF inputs, overlooking calibration errors that cause misalignment and anomaly signals inherent to dToF imaging, limiting real-world applicability. To address these challenges, we propose a novel completion-based method, named DEPTHOR, featuring advances in both the training strategy and model architecture. First, we propose a method to simulate real-world dToF data from the accurate ground truth in synthetic datasets to enable noise-robust training. Second, we design a novel network that incorporates monocular depth estimation (MDE), leveraging global depth relationships and contextual information to improve prediction in challenging regions. On the ZJU-L5 dataset, our training strategy significantly enhances depth completion models, achieving results comparable to depth super-resolution methods, while our model achieves state-of-the-art results, improving Rel and RMSE by 27% and 18%, respectively. On a more challenging set of dToF samples we collected, our method outperforms SOTA methods on preliminary stereo-based GT, improving Rel and RMSE by 23% and 22%, respectively. Our Code is available at https://github.com/ShadowBbBb/Depthor

cs.CV↗

arXiv2025

FrameFusion: Combining Similarity and Importance for Video Token Reduction on Large Vision Language Models

The increasing demand to process long and high-resolution videos significantly burdens Large Vision-Language Models (LVLMs) due to the enormous number of visual tokens. Existing token reduction methods primarily prune tokens based on importance metrics, such as cumulative attention scores. However, even important tokens may exhibit high redundancy caused by similarity among adjacent video frames and repetitive visual elements. To address this limitation, we propose FrameFusion, a novel token reduction approach integrating similarity-based merging with importance-based pruning. We conduct a thorough study on token similarity characteristics, revealing three key insights: (1) spatially corresponding visual tokens between adjacent frames have higher cosine similarities compared to other token pairs; (2) high token similarities prominently decrease in deeper model layers; and (3) token similarity rankings are highly consistent across different layers. Guided by these observations, FrameFusion computes token similarities exclusively between corresponding visual tokens from adjacent frames, applies token merging at initial successive layers followed by pruning in deeper layers, and adopts a cascaded merging strategy to further enhance efficiency. We evaluate FrameFusion comprehensively across six diverse LVLMs, ranging from 2B to 72B parameters, using five video benchmarks encompassing video retrieval, question-answering, and spatial-temporal understanding tasks. Experiments show that FrameFusion reduces visual tokens by 70%, achieving 1.6-3.6x end-to-end speedups, with an average performance impact of less than 3%. Our code is available at: https://github.com/thu-nics/FrameFusion.

cs.CV↗

arXiv2025

Megrez2 Technical Report

We present Megrez2, a novel lightweight and high-performance language model architecture optimized for device native deployment. Megrez2 introduces a novel cross-layer expert sharing mechanism, which significantly reduces total parameter count by reusing expert modules across adjacent transformer layers while maintaining most of the model's capacity. It also incorporates pre-gated routing, enabling memory-efficient expert loading and faster inference. As the first instantiation of the Megrez2 architecture, we introduce the Megrez2-Preview model, which is pre-trained on a 5-trillion-token corpus and further enhanced through supervised fine-tuning and reinforcement learning with verifiable rewards. With only 3B activated and 7.5B stored parameters, Megrez2-Preview demonstrates competitive or superior performance compared to larger models on a wide range of tasks, including language understanding, instruction following, mathematical reasoning, and code generation. These results highlight the effectiveness of the Megrez2 architecture to achieve a balance between accuracy, efficiency, and deployability, making it a strong candidate for real-world, resource-constrained applications.

cs.CL↗

arXiv2025

HoliTracer: Holistic Vectorization of Geographic Objects from Large-Size Remote Sensing Imagery

With the increasing resolution of remote sensing imagery (RSI), large-size RSI has emerged as a vital data source for high-precision vector mapping of geographic objects. Existing methods are typically constrained to processing small image patches, which often leads to the loss of contextual information and produces fragmented vector outputs. To address these, this paper introduces HoliTracer, the first framework designed to holistically extract vectorized geographic objects from large-size RSI. In HoliTracer, we enhance segmentation of large-size RSI using the Context Attention Net (CAN), which employs a local-to-global attention mechanism to capture contextual dependencies. Furthermore, we achieve holistic vectorization through a robust pipeline that leverages the Mask Contour Reformer (MCR) to reconstruct polygons and the Polygon Sequence Tracer (PST) to trace vertices. Extensive experiments on large-size RSI datasets, including buildings, water bodies, and roads, demonstrate that HoliTracer outperforms state-of-the-art methods. Our code and data are available in https://github.com/vvangfaye/HoliTracer.

cs.CV↗

arXiv2025

Aime: Towards Fully-Autonomous Multi-Agent Framework

Multi-Agent Systems (MAS) powered by Large Language Models (LLMs) are emerging as a powerful paradigm for solving complex, multifaceted problems. However, the potential of these systems is often constrained by the prevalent plan-and-execute framework, which suffers from critical limitations: rigid plan execution, static agent capabilities, and inefficient communication. These weaknesses hinder their adaptability and robustness in dynamic environments. This paper introduces Aime, a novel multi-agent framework designed to overcome these challenges through dynamic, reactive planning and execution. Aime replaces the conventional static workflow with a fluid and adaptive architecture. Its core innovations include: (1) a Dynamic Planner that continuously refines the overall strategy based on real-time execution feedback; (2) an Actor Factory that implements Dynamic Actor instantiation, assembling specialized agents on-demand with tailored tools and knowledge; and (3) a centralized Progress Management Module that serves as a single source of truth for coherent, system-wide state awareness. We empirically evaluated Aime on a diverse suite of benchmarks spanning general reasoning (GAIA), software engineering (SWE-bench Verified), and live web navigation (WebVoyager). The results demonstrate that Aime consistently outperforms even highly specialized state-of-the-art agents in their respective domains. Its superior adaptability and task success rate establish Aime as a more resilient and effective foundation for multi-agent collaboration.

cs.AI↗