arXiv subjects

Yu Zhang

Publications and source records attributed to Yu Zhang.

At least 361 records · Page 20Linked to original sources

arXiv2026

SpecFLASH: A Latent-Guided Semi-autoregressive Speculative Decoding Framework for Efficient Multimodal Generation

Large language models and large multimodal models (LLMs and LMMs) deliver strong generative performance but suffer from slow decoding, a problem that becomes more severe when handling visual inputs, whose sequences typically contain many more tokens with lower information density than text. Speculative decoding accelerates LLM inference by letting a compact draft model propose candidate tokens that are selectively accepted by a larger target model, achieving speed-up without degrading quality. However, existing multimodal speculative decoding approaches largely ignore the structural characteristics of visual representations and usually rely on text-only draft models. In this paper, we introduce SpecFLASH, a speculative decoding framework tailored to LMMs that explicitly exploits multimodal structure when designing the draft model. We first mitigate redundancy in visual token sequences with a lightweight, latent-guided token compression module that compacts visual features while preserving semantics, and then leverage the co-occurrence and local correlations of visual entities via a semi-autoregressive decoding scheme that predicts multiple tokens in a single forward pass. Extensive experiments demonstrate that SpecFLASH consistently surpasses prior speculative decoding baselines, achieving up to $2.68\times$ speed-up on video captioning and $2.55\times$ on visual instruction tuning, relative to the original LMM. Our code is available here: https://github.com/ZihuaEvan/FlashSD/.

cs.CV↗

arXiv2026

GPAIR: Gaussian-Kernel-Based Ultrafast 3D Photoacoustic Iterative Reconstruction

Although the iterative reconstruction (IR) algorithm can substantially correct reconstruction artifacts in photoacoustic (PA) computed tomography (PACT), it suffers from long reconstruction times, especially for large-scale three-dimensional (3D) imaging in which IR takes hundreds of seconds to hours. The computing burden severely limits the practical applicability of IR algorithms. In this work, we proposed an ultrafast IR method for 3D PACT, called Gaussian-kernel-based Ultrafast 3D Photoacoustic Iterative Reconstruction (GPAIR), which achieves orders-of-magnitude acceleration in computing. GPAIR transforms traditional spatial grids with continuous isotropic Gaussian kernels. By deriving analytical closed-form expression for pressure waves and implementing powerful GPU-accelerated differentiable Triton operators, GPAIR demonstrates extraordinary ultrafast sub-second reconstruction speed for 3D targets containing 8.4 million voxels in animal experiments. This revolutionary ultrafast image reconstruction enables near-real-time large-scale 3D PA reconstruction, significantly advancing 3D PACT toward clinical applications.

cs.CV↗

arXiv2026

Scalable Quantum Monte Carlo Method for Polariton Chemistry via Mixed Block Sparsity and Tensor Hypercontraction Method

We present a reduced-scaling auxiliary-field quantum Monte Carlo (AFQMC) framework designed for large molecular systems and ensembles, with or without coupling to optical cavities. Our approach leverages the natural block sparsity of Cholesky decomposition (CD) of electron repulsion integrals in molecular ensembles and employs tensor hypercontraction (THC) to efficiently compress low-rank Cholesky blocks. By representing the Cholesky vectors in a mixed format, keeping high-rank blocks in block-sparse form and compressing low-rank blocks with THC, we reduce the scaling of exchange-energy evaluation from quartic to robust cubic in the number of molecular orbitals, while lowering memory from cubic toward quadratic. Benchmark analyses on one-, two-, and three-dimensional molecular ensembles (up to ~1,200 orbitals) show that: a) the number of nonzeros in Cholesky tensors grows linearly with system size across dimensions; b) the average numerical rank increases sublinearly and does not saturate at these sizes; and (c) rank heterogeneity-some blocks nearly full rank and many low rank, naturally motivating the proposed mixed block sparsity and THC scheme for efficient calculation of exchange energy. We demonstrate that the mixed scheme yields cubic CPU-time scaling with favorable prefactors and preserves AFQMC accuracy.

physics.chem-ph↗

arXiv2026

Baseline Method of the Foundation Model Challenge for Ultrasound Image Analysis

Ultrasound (US) imaging exhibits substantial heterogeneity across anatomical structures and acquisition protocols, posing significant challenges to the development of generalizable analysis models. Most existing methods are task-specific, limiting their suitability as clinically deployable foundation models. To address this limitation, the Foundation Model Challenge for Ultrasound Image Analysis (FM\_UIA~2026) introduces a large-scale multi-task benchmark comprising 27 subtasks across segmentation, classification, detection, and regression. In this paper, we present the official baseline for FM\_UIA~2026 based on a unified Multi-Head Multi-Task Learning (MH-MTL) framework that supports all tasks within a single shared network. The model employs an ImageNet-pretrained EfficientNet--B4 backbone for robust feature extraction, combined with a Feature Pyramid Network (FPN) to capture multi-scale contextual information. A task-specific routing strategy enables global tasks to leverage high-level semantic features, while dense prediction tasks exploit spatially detailed FPN representations. Training incorporates a composite loss with task-adaptive learning rate scaling and a cosine annealing schedule. Validation results demonstrate the feasibility and robustness of this unified design, establishing a strong and extensible baseline for ultrasound foundation model research. The code and dataset are publicly available at \href{https://github.com/lijiake2408/Foundation-Model-Challenge-for-Ultrasound-Image-Analysis}{GitHub}.

cs.CV↗

arXiv2026

Quantum-Enhanced Sensing Enabled by Scrambling-Induced Genuine Multipartite Entanglement

Quantum sensing leverages quantum resources to surpass the standard quantum limit, yet many existing protocols rely on the preparation of complex entangled states and Hamiltonian engineering, posing challenges for universality and scalability. Here, we report an experimental realization of a universal protocol, known as Butterfly Metrology, proposed in [arXiv:2411.12794], demonstrating a scrambling-based approach for quantum-enhanced sensing on a superconducting quantum processor. By exploiting many-body information scrambling, we observe quantum-enhanced sensitivity to an encoded phase beyond the standard quantum limit, with a scaling consistent with a factor-of-two of the Heisenberg limit for system sizes of up to 10 qubits. Importantly, we experimentally establish a connection between the enhanced sensitivity and the dynamics of the out-of-time-order correlator (OTOC), and show that the buildup of scrambling-induced genuine multipartite entanglement underlies the observed sensitivity enhancement. Our results demonstrate a scalable and practical approach for quantum-enhanced sensing in interacting many-body quantum systems.

quant-ph↗

arXiv2026

Masked Diffusion Generative Recommendation

Generative recommendation (GR) typically first quantizes continuous item embeddings into multi-level semantic IDs (SIDs), and then generates the next item via autoregressive decoding. Although existing methods are already competitive in terms of recommendation performance, directly inheriting the autoregressive decoding paradigm from language models still suffers from three key limitations: (1) autoregressive decoding struggles to jointly capture global dependencies among the multi-dimensional features associated with different positions of SID; (2) using a unified, fixed decoding path for the same item implicitly assumes that all users attend to item attributes in the same order; (3) autoregressive decoding is inefficient at inference time and struggles to meet real-time requirements. To tackle these challenges, we propose MDGR, a Masked Diffusion Generative Recommendation framework that reshapes the GR pipeline from three perspectives: codebook, training, and inference. (1) We adopt a parallel codebook to provide a structural foundation for diffusion-based GR. (2) During training, we adaptively construct masking supervision signals along both the temporal and sample dimensions. (3) During inference, we develop a warm-up-based two-stage parallel decoding strategy for efficient generation of SIDs. Extensive experiments on multiple public and industrial-scale datasets show that MDGR outperforms ten state-of-the-art baselines by up to 10.78%. Furthermore, by deploying MDGR on a large-scale online advertising platform, we achieve a 1.20% increase in revenue, demonstrating its practical value.

cs.IR↗

arXiv2026

Resilient Grid Hardening against Multiple Hazards: An Adaptive Two-Stage Stochastic Optimization Approach

The growing prevalence of extreme weather events driven by climate change poses significant challenges to power system resilience. Infrastructure damage and prolonged power outages highlight the urgent need for effective grid-hardening strategies. While some measures provide long-term protection against specific hazards, they can become counterproductive under conflicting threats. In this work, we develop an adaptive two-stage stochastic optimization framework to support dynamic decision-making for hardening critical grid components under multiple hazard exposures. Unlike traditional approaches, our model adapts to evolving climate conditions, enabling more resilient investment strategies. Furthermore, we integrate long-term (undergrounding) and short-term (vegetation management) hardening actions to jointly minimize total system costs. Extensive simulation results validate the effectiveness of the proposed framework in reducing outage and repair costs while enhancing the adaptability and robustness of grid infrastructure planning.

eess.SY↗

arXiv2026

Scattering theory of frequency-entangled biphoton states facilitated by cavity polaritons

The use of quantum light to probe exciton properties in semiconductor and molecular nanostructures typically occurs in the low-intensity regime. A substantial enhancement of exciton-photon coupling can be achieved with photonic cavities, where excitons hybridize with cavity modes to form polariton states. To provide a theoretical framework for interpreting emerging experimental efforts in this direction, we develop a scattering theory describing the interaction of frequency-entangled photon pairs with cavity polariton and bipolariton states under various coupling regimes. Employing the Tavis-Cummings model in combination with our scattering approach, we present a quantitative analysis of how the interaction of the entangled photon pair with the polariton/bipolariton modifies its joint spectral amplitude (JSA). Specifically, we examine the effects of the cavity-mode steady-state population, exciton-cavity coupling strength, and different forms of the input photon JSA. Our results show that the entanglement entropy of the scattered photons is highly sensitive to the interplay between the input JSA and the spectral line shapes of the polariton resonances, emphasizing the cavity filtering effects. We suggest that biphoton scattering quantum light spectroscopy best serves as a sensitive probe of polariton and bipolariton states in the photon-vacuum cavity state. Our approach is not only robust to various regimes of cavity-exciton coupling, but also amenable to extensions beyond the Tavis-Cummings model, enabling the representation of a broad class of molecular systems and solid state quantum materials.

quant-ph↗

arXiv2026

Cohesive Group Discovery in Interaction Graphs under Explicit Density Constraints

Discovering cohesive groups is a fundamental primitive in graph-based recommender systems, underpinning tasks such as social recommendation, bundle discovery, and community-aware modeling. In interaction graphs, cohesion is often modeled as the $γ$-quasi-clique, an induced subgraph whose internal edge density meets a user-defined threshold $γ$. This formulation provides explicit control over within-group connectivity while accommodating the sparsity inherent in real-world data. This paper presents EDQC, an effective framework for cohesive group discovery under explicit density constraints. EDQC leverages a lightweight energy diffusion process to rank vertices for localizing promising candidate regions. Guided by this ranking, the framework extracts and refines a candidate subgraph to ensure the output strictly satisfies the target density requirement. Extensive experiments on 75 real-world graphs across varying density thresholds demonstrate that EDQC identifies the largest mean $γ$-quasi-cliques in the vast majority of cases, achieving lower variance than the state-of-the-art methods while maintaining competitive runtime. Furthermore, statistical analysis confirms that EDQC significantly outperforms the baselines, underscoring its robustness and practical utility for cohesive group discovery in graph-based recommender systems.

cs.SI↗

arXiv2026

Synthetic Singers: A Review of Deep-Learning-based Singing Voice Synthesis Approaches

Recent advances in singing voice synthesis (SVS) have attracted substantial attention from both academia and industry. With the advent of large language models and novel generative paradigms, producing controllable, high-fidelity singing voices has become an attainable goal. Yet the field still lacks a comprehensive survey that systematically analyzes deep-learning-based singing voice synthesis systems and their enabling technologies. To address the aforementioned issue, this survey first categorizes existing systems by task type and then organizes current architectures into two major paradigms: cascaded and end-to-end approaches. Moreover, we provide an in-depth analysis of core technologies, covering singing modeling and control techniques. Finally, we review relevant datasets, annotation tools, and evaluation benchmarks that support training and assessment. In appendix, we introduce training strategies and further discussion of SVS. This survey provides an up-to-date review of the literature on SVS models, which would be a useful reference for both researchers and engineers. Related materials are available at https://github.com/David-Pigeon/SyntheticSingers.

eess.AS↗

arXiv2026

Surrogate Modeling via Factorization Machine and Ising Model with Enhanced Higher-Order Interaction Learning

Recently, a surrogate model was proposed that employs a factorization machine to approximate the underlying input-output mapping of the original system, with quantum annealing used to optimize the resulting surrogate function. Inspired by this approach, we propose an enhanced surrogate model that incorporates additional slack variables into both the factorization machine and its associated Ising representation thereby unifying what was by design a two-step process into a single, integrated step. During the training phase, the slack variables are iteratively updated, enabling the model to account for higher-order feature interactions. We apply the proposed method to the task of predicting drug combination effects. Experimental results indicate that the introduction of slack variables leads to a notable improvement of performance. Our algorithm offers a promising approach for building efficient surrogate models that exploit potential quantum advantages.

cs.LG↗

arXiv2026

Holos: A Web-Scale LLM-Based Multi-Agent System for the Agentic Web

As large language models (LLM)-driven agents transition from isolated task solvers to persistent digital entities, the emergence of the Agentic Web, an ecosystem where heterogeneous agents autonomously interact and co-evolve, marks a pivotal shift toward Artificial General Intelligence (AGI). However, LLM-based multi-agent systems (LaMAS) are hindered by open-world issues such as scaling friction, coordination breakdown, and value dissipation. To address these challenges, we introduce Holos, a web-scale LaMAS architected for long-term ecological persistence. Holos adopts a five-layer architecture, with core modules primarily featuring the Nuwa engine for high-efficiency agent generation and hosting, a market-driven Orchestrator for resilient coordination, and an endogenous value cycle to achieve incentive compatibility. By bridging the gap between micro-level collaboration and macro-scale emergence, Holos hopes to lay the foundation for the next generation of the self-organizing and continuously evolving Agentic Web. We have publicly released Holos (accessible at https://holosai.io), providing a resource for the community and a testbed for future research in large-scale agentic ecosystems.

cs.AI↗

arXiv2026

ZPD Detector: Data Selection via Capability-Difficulty Alignment for Large Language Models

As the cost of training large language models continues to increase and high-quality training data become increasingly scarce, selecting high-value samples or synthesizing effective training data under limited data budgets has emerged as a critical research problem. Most existing data selection methods rely on static criteria, such as difficulty, uncertainty, or heuristics, and fail to model the evolving relationship between the model and the data. Inspired by the educational theory of the Zone of Proximal Development (ZPD), we propose ZPD Detector, a data selection framework that adopts a bidirectional perspective between models and data by explicitly modeling the alignment between sample difficulty and the model's current capability. ZPD Detector integrates difficulty calibration, model capability estimation based on Item Response Theory (IRT), and a capability-difficulty matching score to dynamically identify the most informative samples at each learning stage, improving data utilization efficiency; moreover, this dynamic matching strategy provides new insights into training strategy design. All code and data will be released after our work be accepted to support reproducible researc

cs.CL↗

arXiv2026

Knowledge Homophily in Large Language Models

Large Language Models (LLMs) have been increasingly studied as neural knowledge bases for supporting knowledge-intensive applications such as question answering and fact checking. However, the structural organization of their knowledge remains unexplored. Inspired by cognitive neuroscience findings, such as semantic clustering and priming, where knowing one fact increases the likelihood of recalling related facts, we investigate an analogous knowledge homophily pattern in LLMs. To this end, we map LLM knowledge into a graph representation through knowledge checking at both the triplet and entity levels. After that, we analyze the knowledgeability relationship between an entity and its neighbors, discovering that LLMs tend to possess a similar level of knowledge about entities positioned closer in the graph. Motivated by this homophily principle, we propose a Graph Neural Network (GNN) regression model to estimate entity-level knowledgeability scores for triplets by leveraging their neighborhood scores. The predicted knowledgeability enables us to prioritize checking less well-known triplets, thereby maximizing knowledge coverage under the same labeling budget. This not only improves the efficiency of active labeling for fine-tuning to inject knowledge into LLMs but also enhances multi-hop path retrieval in reasoning-intensive question answering.

cs.LG↗

arXiv2026

AgriAgent: Contract-Driven Planning and Capability-Aware Tool Orchestration in Real-World Agriculture

Intelligent agent systems in real-world agricultural scenarios must handle diverse tasks under multimodal inputs, ranging from lightweight information understanding to complex multi-step execution. However, most existing approaches rely on a unified execution paradigm, which struggles to accommodate large variations in task complexity and incomplete tool availability commonly observed in agricultural environments. To address this challenge, we propose AgriAgent, a two-level agent framework for real-world agriculture. AgriAgent adopts a hierarchical execution strategy based on task complexity: simple tasks are handled through direct reasoning by modality-specific agents, while complex tasks trigger a contract-driven planning mechanism that formulates tasks as capability requirements and performs capability-aware tool orchestration and dynamic tool generation, enabling multi-step and verifiable execution with failure recovery. Experimental results show that AgriAgent achieves higher execution success rates and robustness on complex tasks compared to existing tool-centric agent baselines that rely on unified execution paradigms. All code, data will be released at after our work be accepted to promote reproducible research.

cs.CL↗

arXiv2026

Analytical Expression for Spherically Symmetric Photoacoustic Sources: A Unified General Solution (Theoretical Analysis and Derivation)

Here we present a comprehensive derivation of the analytical expression for the spatiotemporal acoustic pressure generated by photoacoustic sources with spherically symmetric initial pressure distributions. Starting from the fundamental photoacoustic wave equation, we derive a unified analytical solution applicable to arbitrary spherically symmetric initial distributions. Specific expressions are provided for several common distributions including uniform spherical sources, Gaussian distributions, exponential distributions, and power-law distributions. Far-field approximations are also discussed. The derived expressions provide valuable tools for photoacoustic imaging system design and signal analysis. We provide codes for ultrafast forward simulation using the general analytical spherically symmetric model, the implementation is available in the GitHub repository: \href{https://github.com/JaegerCQ/SlingBAG_Ultra}.

physics.optics↗

arXiv2026

Search for Ultralight Axion Dark Matter with a Levitated Ferromagnetic Torsional Oscillator

We present a search for ultralight axion dark matter coupled to electron spins using a levitated ferromagnetic torsional oscillator (FMTO). This platform directly measures axion-induced torques on a macroscopic spin-polarized body, combining large spin density with strong mechanical isolation to probe magnetic fluctuations below 10 Hz while suppressing gradient-field noise. In a first implementation, the experiment yielded 18000 s of analyzable data at room temperature under high vacuum with optical readout and triple-layer magnetic shielding. A likelihood-based statistical framework, incorporating stochastic fluctuations in the axion-field amplitude, was used to evaluate the data. No excess consistent with an axion-induced pseudo-magnetic field was observed near 2e-14 eV. To account for possible shielding-induced signal attenuation, we quantify its effect and report both the uncorrected (g_aee < 1e-7) and attenuation-corrected (g_aee < 6e-5) 90% CL limits on the axion-electron coupling. Looking ahead, improvements guided by both noise-budget analysis and shielding-attenuation considerations, including optimized levitation geometry, cryogenic operation, and superconducting shielding, are expected to boost sensitivity by multiple orders of magnitude.

hep-ph↗

arXiv2026

SlingBAG Pro: Accelerating point cloud-based iterative reconstruction for 3D photoacoustic imaging with arbitrary array geometries

High-quality three-dimensional (3D) photoacoustic imaging (PAI) is gaining increasing attention in clinical applications. To address the challenges of limited space and high costs, irregular geometric transducer arrays that conform to specific imaging regions are promising for achieving high-quality 3D PAI with fewer transducers. However, traditional iterative reconstruction algorithms struggle with irregular array configurations, suffering from high computational complexity, substantial memory requirements, and lengthy reconstruction times. In this work, we introduce SlingBAG Pro, an advanced reconstruction algorithm based on the point cloud iteration concept of the Sliding ball adaptive growth (SlingBAG) method, while extending its compatibility to arbitrary array geometries. SlingBAG Pro maintains high reconstruction quality, reduces the number of required transducers, and employs a hierarchical optimization strategy that combines zero-gradient filtering with progressively increased temporal sampling rates during iteration. This strategy rapidly removes redundant spatial point clouds, accelerates convergence, and significantly shortens overall reconstruction time. Compared to the original SlingBAG algorithm, SlingBAG Pro achieves up to a 2.2-fold speed improvement in point cloud-based 3D PA reconstruction under irregular array geometries. The proposed method is validated through both simulation and in vivo mouse experiments, and the source code is publicly available at https://github.com/JaegerCQ/SlingBAG_Pro.

cs.CV↗