arXiv subjects

Yukun Zhang

Publications and source records attributed to Yukun Zhang.

At least 55 records · Page 3Linked to original sources

arXiv2025

Beyond Private or Public: Large Language Models as Quasi-Public Goods in the AI Economy

This paper conceptualizes Large Language Models (LLMs) as a form of mixed public goods within digital infrastructure, analyzing their economic properties through a comprehensive theoretical framework. We develop mathematical models to quantify the non-rivalry characteristics, partial excludability, and positive externalities of LLMs. Through comparative analysis of open-source and closed-source development paths, we identify systematic differences in resource allocation efficiency, innovation trajectories, and access equity. Our empirical research evaluates the spillover effects and network externalities of LLMs across different domains, including knowledge diffusion, innovation acceleration, and industry transformation. Based on these findings, we propose policy recommendations for balancing innovation incentives with equitable access, including public-private partnership mechanisms, computational resource democratization, and governance structures that optimize social welfare. This interdisciplinary approach contributes to understanding the economic nature of foundation AI models and provides policy guidance for their development as critical digital infrastructure

cs.CY↗

arXiv2025

Heisenberg limited quantum algorithm for estimating the fidelity susceptibility

The fidelity susceptibility serves as a universal probe for quantum phase transitions, offering an order-parameter-free metric that captures ground-state sensitivity to Hamiltonian perturbations and exhibits critical scaling. Classical computation of this quantity, however, is limited by exponential Hilbert space growth and correlation divergence near criticality, restricting analyses to small or specialized systems. Here, we present a quantum algorithm that achieves efficient and Heisenberg-limited estimation of fidelity susceptibility through a novel resolvent reformulation, leveraging quantum singular value transformation for pseudoinverse block encoding with amplitude estimation for norm evaluation. This constitutes the first quantum algorithm for fidelity susceptibility with optimal precision scaling. Moreover, for frustration-free Hamiltonians, we show that the resolvent can be approximated with a further quadratic speedup. Our work bridges quantum many-body physics and algorithmic design, enabling scalable exploration of quantum criticality with applications in materials simulation, metrology, and beyond on fault-tolerant quantum platforms.

quant-ph↗

arXiv2025

Measuring Less to Learn More: Quadratic Speedup in learning Nonlinear Properties of Quantum Density Matrices

A fundamental task in quantum information science is to measure nonlinear functionals of quantum states, such as $\mathrm{Tr}(ρ^k O)$. Intuitively, one expects that computing a $k$-th order quantity generally requires $O(k)$ copies of the state $ρ$, and we rigorously establish this lower bound under sample access to $ρ$. Surprisingly, this limitation can be overcome when one has purified access via a unitary that prepares a purification of $ρ$, a scenario naturally arising in quantum simulation and computation. In this setting, we find a different lower bound of $Θ(\sqrt{k})$, and present a quantum algorithm that achieves this bound, demonstrating a quadratic advantage over sample-based methods. The key technical innovation lies in a designed quantum algorithm and optimal polynomial approximation theory -- specifically, Chebyshev polynomial approximations tailored to the boundary behavior of power functions. Our results unveil a fundamental distinction between sample and purified access to quantum states, with broad implications for estimating quantum entropies and quantum Fisher information, realizing quantum virtual distillation and cooling, and evaluating other multiple nonlinear quantum observables with classical shadows.

quant-ph↗

arXiv2025

Continuous-Time Attention: PDE-Guided Mechanisms for Long-Sequence Transformers

We propose a novel framework, Continuous_Time Attention, which infuses partial differential equations (PDEs) into the Transformer's attention mechanism to address the challenges of extremely long input sequences. Instead of relying solely on a static attention matrix, we allow attention weights to evolve over a pseudo_time dimension via diffusion, wave, or reaction_diffusion dynamics. This mechanism systematically smooths local noise, enhances long_range dependencies, and stabilizes gradient flow. Theoretically, our analysis shows that PDE_based attention leads to better optimization landscapes and polynomial rather than exponential decay of distant interactions. Empirically, we benchmark our method on diverse experiments_demonstrating consistent gains over both standard and specialized long sequence Transformer variants. Our findings highlight the potential of PDE_based formulations to enrich attention mechanisms with continuous_time dynamics and global coherence.

cs.LG↗

arXiv2025

Few-shot Semantic Encoding and Decoding for Video Surveillance

With the continuous increase in the number and resolution of video surveillance cameras, the burden of transmitting and storing surveillance video is growing. Traditional communication methods based on Shannon's theory are facing optimization bottlenecks. Semantic communication, as an emerging communication method, is expected to break through this bottleneck and reduce the storage and transmission consumption of video. Existing semantic decoding methods often require many samples to train the neural network for each scene, which is time-consuming and labor-intensive. In this study, a semantic encoding and decoding method for surveillance video is proposed. First, the sketch was extracted as semantic information, and a sketch compression method was proposed to reduce the bit rate of semantic information. Then, an image translation network was proposed to translate the sketch into a video frame with a reference frame. Finally, a few-shot sketch decoding network was proposed to reconstruct video from sketch. Experimental results showed that the proposed method achieved significantly better video reconstruction performance than baseline methods. The sketch compression method could effectively reduce the storage and transmission consumption of semantic information with little compromise on video quality. The proposed method provides a novel semantic encoding and decoding method that only needs a few training samples for each surveillance scene, thus improving the practicality of the semantic communication system.

cs.CV↗

arXiv2025

NTIRE 2025 Challenge on Event-Based Image Deblurring: Methods and Results

This paper presents an overview of NTIRE 2025 the First Challenge on Event-Based Image Deblurring, detailing the proposed methodologies and corresponding results. The primary goal of the challenge is to design an event-based method that achieves high-quality image deblurring, with performance quantitatively assessed using Peak Signal-to-Noise Ratio (PSNR). Notably, there are no restrictions on computational complexity or model size. The task focuses on leveraging both events and images as inputs for single-image deblurring. A total of 199 participants registered, among whom 15 teams successfully submitted valid results, offering valuable insights into the current state of event-based image deblurring. We anticipate that this challenge will drive further advancements in event-based vision research.

cs.CV↗

arXiv2025

Hamiltonian Dynamics Learning: A Scalable Approach to Quantum Process Characterization

Quantum process characterization is a fundamental task in quantum information processing, yet conventional methods, such as quantum process tomography, require prohibitive resources and lack scalability. Here, we introduce an efficient quantum process learning method specifically designed for short-time Hamiltonian dynamics. Our approach reconstructs an equivalent quantum circuit representation from measurement data of unknown Hamiltonian evolution without requiring additional assumptions and achieves polynomial sample and computational efficiency. Our results have broad applications in various directions. We demonstrate applications in quantum machine learning, where our protocol enables efficient training of variational quantum neural networks by directly learning unitary transformations. Additionally, it facilitates the prediction of quantum expectation values with provable efficiency and provides a robust framework for verifying quantum computations and benchmarking realistic noisy quantum hardware. This work establishes a new theoretical foundation for practical quantum dynamics learning, paving the way for scalable quantum process characterization in both near-term and fault-tolerant quantum computing.

quant-ph↗

arXiv2025

Unveiling LLM Mechanisms Through Neural ODEs and Control Theory

This paper proposes a framework combining Neural Ordinary Differential Equations (Neural ODEs) and robust control theory to enhance the interpretability and control of large language models (LLMs). By utilizing Neural ODEs to model the dynamic evolution of input-output relationships and introducing control mechanisms to optimize output quality, we demonstrate the effectiveness of this approach across multiple question-answer datasets. Experimental results show that the integration of Neural ODEs and control theory significantly improves output consistency and model interpretability, advancing the development of explainable AI technologies.

cs.LG↗

arXiv2025

AI-Enabled Rent-Seeking: How Generative AI Alters Market Transparency and Efficiency

The rapid advancement of generative artificial intelligence (AI) has transformed the information environment, creating both opportunities and challenges. This paper explores how generative AI influences economic rent-seeking behavior and its broader impact on social welfare. We develop a dynamic economic model involving multiple agents who may engage in rent-seeking activities and a regulator aiming to mitigate social welfare losses. Our analysis reveals a dual effect of generative AI: while it reduces traditional information rents by increasing transparency, it also introduces new forms of rent-seeking, such as information manipulation and algorithmic interference. These behaviors can lead to decreased social welfare by exacerbating information asymmetries and misallocating resources. To address these challenges, we propose policy interventions, including taxation and regulatory measures. This study provides a new perspective on the economic implications of generative AI, offering valuable insights for policymakers and laying a foundation for future research on regulating AI-driven economic behaviors.

cs.CY↗

arXiv2025

Generative AI and Information Asymmetry: Impacts on Adverse Selection and Moral Hazard

Information asymmetry often leads to adverse selection and moral hazard in economic markets, causing inefficiencies and welfare losses. Traditional methods to address these issues, such as signaling and screening, are frequently insufficient. This research investigates how Generative Artificial Intelligence (AI) can create detailed informational signals that help principals better understand agents' types and monitor their actions. By incorporating these AI-generated signals into a principal-agent model, the study aims to reduce inefficiencies and improve contract designs. Through theoretical analysis and simulations, we demonstrate that Generative AI can effectively mitigate adverse selection and moral hazard, resulting in more efficient market outcomes and increased social welfare. Additionally, the findings offer practical insights for policymakers and industry stakeholders on the responsible implementation of Generative AI solutions to enhance market performance.

cs.CY↗

arXiv2025

Exploring the Head Effect in Live Streaming Platforms: A Two-Sided Market and Welfare Analysis

We develop a comprehensive theoretical framework to analyze live streaming platforms as two-sided markets, focusing on the head effect where a small subset of elite streamers disproportionately attracts viewer attention. By constructing both static and dynamic models, we capture the interplay between network effects, content quality investments, and platform policies-such as commission structures and traffic allocation algorithms-that drive traffic concentration. Our welfare analysis demonstrates that although short-term consumer utility may benefit from concentrated viewership, long-term content diversity and overall social welfare are adversely impacted. Extensive simulations further validate our models and show that targeted policy interventions can rebalance viewer distribution and mitigate winner-takes-all dynamics. These findings offer actionable insights for platform designers and regulators in the digital economy.

cs.CY↗

arXiv2025

Revenue vs. Welfare: A Comprehensive Analysis of Strategic Trade-offs in Online Food Delivery Systems

This paper investigates the trade-off between short-term revenue generation and long-term social welfare optimization in online food delivery platforms. We first develop a static model that captures the equilibrium interactions among restaurants, consumers, and delivery workers, using Gross Merchandise Value (GMV) as a proxy for immediate performance. Building on this, we extend our analysis to a dynamic model that integrates evolving state variables,such as platform reputation and participant retention-to capture long-term behavior. By applying dynamic programming techniques, we derive optimal strategies that balance GMV maximization with social welfare enhancement. Extensive multi-agent simulations validate our theoretical predictions, demonstrating that while a GMV-focused approach yields strong initial gains, it ultimately undermines long-term stability. In contrast, a social welfare-oriented strategy produces more sustainable and robust outcomes. Our findings provide actionable insights for platform operators and policymakers seeking to harmonize rapid growth with long-term

cs.GT↗

arXiv2025

Fault-tolerant quantum algorithms for quantum molecular systems: A survey

Solving quantum molecular systems presents a significant challenge for classical computation. The advent of early fault-tolerant quantum computing (EFTQC) devices offers a promising avenue to address these challenges, leveraging advanced quantum algorithms with reduced hardware requirements. This review surveys the latest developments in EFTQC and fully fault-tolerant quantum computing (FFTQC) algorithms for quantum molecular systems, covering encoding schemes, advanced Hamiltonian simulation techniques, and ground-state energy estimation methods. We highlight recent progress in overcoming practical barriers, such as reducing circuit depth and minimizing the use of ancillary qubits. Special attention is given to the potential quantum advantages achievable through these algorithms, as well as the limitations imposed by dequantization and classical simulation techniques. The review concludes with a discussion of future directions, emphasizing the need for optimized algorithms and experimental validation to bridge the gap between theoretical developments and practical implementation in EFTQC and FFTQC for quantum molecular systems.

quant-ph↗

arXiv2025

ToolFactory: Automating Tool Generation by Leveraging LLM to Understand REST API Documentations

LLM-based tool agents offer natural language interfaces, enabling users to seamlessly interact with computing services. While REST APIs are valuable resources for building such agents, they must first be transformed into AI-compatible tools. Automatically generating AI-compatible tools from REST API documents can greatly streamline tool agent development and minimize user learning curves. However, API documentation often suffers from a lack of standardization, inconsistent schemas, and incomplete information. To address these issues, we developed \textbf{ToolFactory}, an open-source pipeline for automating tool generation from unstructured API documents. To enhance the reliability of the developed tools, we implemented an evaluation method to diagnose errors. Furthermore, we built a knowledge base of verified tools, which we leveraged to infer missing information from poorly documented APIs. We developed the API Extraction Benchmark, comprising 167 API documents and 744 endpoints in various formats, and designed a JSON schema to annotate them. This annotated dataset was utilized to train and validate ToolFactory. The experimental results highlight the effectiveness of ToolFactory. We also demonstrated ToolFactory by creating a domain-specific AI agent for glycomaterials research. ToolFactory exhibits significant potential for facilitating the seamless integration of scientific REST APIs into AI workflows.

cs.LG↗

arXiv2024

A Dequantized Algorithm for the Guided Local Hamiltonian Problem

The local Hamiltonian (LH) problem, the quantum analog of the classical constraint satisfaction problem, is a cornerstone of quantum computation and complexity theory. It is known to be QMA-complete, indicating that it is challenging even for quantum computers. Interestingly, the guided local Hamiltonian (GLH) problem -- an LH problem with a guiding state that has a non-trivial overlap with the ground state -- can be efficiently solved on a quantum computer and is proved to be BQP-complete. This makes the GLH problem a valuable framework for exploring the fundamental separation between classical and quantum computation. Remarkably, the quantum algorithm for solving the GLH problem can be `dequantized' (i.e., made classically simulatable) under certain conditions, such as when only constant accuracy is required and when the Hamiltonian satisfies an unrealistic constant operator norm constraint. In this work, we relieve these restrictions by introducing a dequantized classical algorithm for a randomized quantum imaginary-time evolution quantum algorithm. We demonstrate that it achieves either limited or arbitrary constant accuracy, depending on whether the guiding state's overlap is general or exceeds a certain threshold. Crucially, our approach eliminates the constant operator norm constraint on the Hamiltonian, opening its applicability to realistic problems. Our results advance the classical solution of the GLH problem in practical settings and provide new insights into the boundary between classical and quantum computational power.

quant-ph↗

arXiv2024

FedCVD: The First Real-World Federated Learning Benchmark on Cardiovascular Disease Data

Cardiovascular diseases (CVDs) are currently the leading cause of death worldwide, highlighting the critical need for early diagnosis and treatment. Machine learning (ML) methods can help diagnose CVDs early, but their performance relies on access to substantial data with high quality. However, the sensitive nature of healthcare data often restricts individual clinical institutions from sharing data to train sufficiently generalized and unbiased ML models. Federated Learning (FL) is an emerging approach, which offers a promising solution by enabling collaborative model training across multiple participants without compromising the privacy of the individual data owners. However, to the best of our knowledge, there has been limited prior research applying FL to the cardiovascular disease domain. Moreover, existing FL benchmarks and datasets are typically simulated and may fall short of replicating the complexity of natural heterogeneity found in realistic datasets that challenges current FL algorithms. To address these gaps, this paper presents the first real-world FL benchmark for cardiovascular disease detection, named FedCVD. This benchmark comprises two major tasks: electrocardiogram (ECG) classification and echocardiogram (ECHO) segmentation, based on naturally scattered datasets constructed from the CVD data of seven institutions. Our extensive experiments on these datasets reveal that FL faces new challenges with real-world non-IID and long-tail data. The code and datasets of FedCVD are available https://github.com/SMILELab-FL/FedCVD.

eess.SP↗

arXiv2024

Unraveling Text Generation in LLMs: A Stochastic Differential Equation Approach

This paper explores the application of Stochastic Differential Equations (SDE) to interpret the text generation process of Large Language Models (LLMs) such as GPT-4. Text generation in LLMs is modeled as a stochastic process where each step depends on previously generated content and model parameters, sampling the next word from a vocabulary distribution. We represent this generation process using SDE to capture both deterministic trends and stochastic perturbations. The drift term describes the deterministic trends in the generation process, while the diffusion term captures the stochastic variations. We fit these functions using neural networks and validate the model on real-world text corpora. Through numerical simulations and comprehensive analyses, including drift and diffusion analysis, stochastic process property evaluation, and phase space exploration, we provide deep insights into the dynamics of text generation. This approach not only enhances the understanding of the inner workings of LLMs but also offers a novel mathematical perspective on language generation, which is crucial for diagnosing, optimizing, and controlling the quality of generated text.

cs.LG↗

arXiv2024

Joint Optimization of Buffer Delay and HARQ for Video Communications

To improve the quality of experience (QoE) in video communication over lossy networks, this paper presents a transmission method that jointly optimizes buffer delay and Hybrid Automatic Repeat request (HARQ), referred to as BD-HARQ. This method operates on packet group and employs dynamic buffer delay combined with HARQ strategy for transmission. By defining the QoE based on metrics such as buffer delay, Forward Error Correction (FEC) redundancy, and data recovery rate, the proposed method derives its closed-form expression through rigorous mathematical modeling and analysis. The optimal transmission parameters, i.e., the buffer delay and the FEC redundancy, are then determined and implemented, guaranteeing the real-time performance, transmission efficiency, and data recovery rate of video communication. Experimental results demonstrate that the proposed method aligns well with its theoretical expectations, and that it can provide up to 13.7% higher QoE compared to existing methods and increase the tolerance for packet loss rate from 15%-22% to up to 31% while maintaining a high QoE.

cs.MM↗