arXiv subjects

Minghui Xu

Publications and source records attributed to Minghui Xu.

At least 55 records · Page 3Linked to original sources

arXiv2025

xRWA: A Cross-Chain Framework for Interoperability of Real-World Assets

Real-World Assets (RWAs) have recently attracted increasing attention as a means of bridging traditional financial instruments with decentralized infrastructures. By representing assets such as bonds, commodities, and real estate on blockchains, RWAs can enhance liquidity, broaden accessibility, and extend the scope of decentralized finance. Industry forecasts further suggest rapid growth of tokenized RWAs in the coming years, underscoring their potential role in the evolution of digital financial markets. However, when deployed across multiple blockchains, RWAs face challenges such as repeated authentication on different chains and inefficiency caused by multi-step settlement protocols. To address these issues, we present a cross-chain framework for RWAs that emphasizes identity management, authentication, and interaction. The framework integrates Decentralized Identifiers and Verifiable Credentials with customized attributes to support decentralized identification, and incorporates an authentication protocol based on Simplified Payment Verification to avoid redundant verification across chains. Furthermore, we design a cross-chain channel that enables the settlement of RWAs without requiring channel closure, thereby improving operational efficiency. We implement the framework and evaluate it through simulations, which confirm its feasibility and demonstrate improvements in efficiency for RWAs in cross-chain settings.

cs.CR↗

arXiv2025

PIR-DSN: A Decentralized Storage Network Supporting Private Information Retrieval

Decentralized Storage Networks (DSNs) are emerging as a foundational infrastructure for Web 3.0, offering global peer-to-peer storage. However, a critical vulnerability persists: user privacy during file retrieval remains largely unaddressed, risking the exposure of sensitive information. To overcome this, we introduce PIR-DSN, the first DSN protocol to integrate Private Information Retrieval (PIR) for both single and multi-server settings. Our key innovations include a novel secure mapping method that transforms sparse file identifiers into compact integer indexes, enabling both public verifiability of file operations and efficient private retrieval. Furthermore, PIR-DSN guarantees Byzantine-robust private retrieval through file replication across multiple miners. We implement and rigorously evaluate PIR-DSN against three prominent industrial DSN systems. Experimental results demonstrate that PIR-DSN achieves comparable overhead for file upload and deletion. While PIR inherently introduces an additional computational cost leading to higher retrieval latency, PIR-DSN maintains comparable throughput. These findings underscore PIR-DSN's practical viability for privacy-sensitive applications within DSN environments.

cs.DC↗

arXiv2025

Beyond Model Jailbreak: Systematic Dissection of the "Ten DeadlySins" in Embodied Intelligence

Embodied AI systems integrate language models with real world sensing, mobility, and cloud connected mobile apps. Yet while model jailbreaks have drawn significant attention, the broader system stack of embodied intelligence remains largely unexplored. In this work, we conduct the first holistic security analysis of the Unitree Go2 platform and uncover ten cross layer vulnerabilities the "Ten Sins of Embodied AI Security." Using BLE sniffing, traffic interception, APK reverse engineering, cloud API testing, and hardware probing, we identify systemic weaknesses across three architectural layers: wireless provisioning, core modules, and external interfaces. These include hard coded keys, predictable handshake tokens, WiFi credential leakage, missing TLS validation, static SSH password, multilingual safety bypass behavior, insecure local relay channels, weak binding logic, and unrestricted firmware access. Together, they allow adversaries to hijack devices, inject arbitrary commands, extract sensitive information, or gain full physical control.Our findings show that securing embodied AI requires far more than aligning the model itself. We conclude with system level lessons learned and recommendations for building embodied platforms that remain robust across their entire software hardware ecosystem.

cs.CR↗

arXiv2025

"MCP Does Not Stand for Misuse Cryptography Protocol": Uncovering Cryptographic Misuse in Model Context Protocol at Scale

The Model Context Protocol (MCP) is rapidly emerging as the middleware for LLM-based applications, offering a standardized interface for tool integration. However, its built-in security mechanisms are minimal: while schemas and declarations prevent malformed requests, MCP provides no guarantees of authenticity or confidentiality, forcing developers to implement cryptography themselves. Such ad hoc practices are historically prone to misuse, and within MCP they threaten sensitive data and services. We present MICRYSCOPE, the first domain-specific framework for detecting cryptographic misuses in MCP implementations. MICRYSCOPE combines three key innovations: a cross-language intermediate representation that normalizes cryptographic APIs across diverse ecosystems, a hybrid dependency analysis that uncovers explicit and implicit function relationships (including insecure runtime compositions orchestrated by LLMs) and a taint-based misuse detector that tracks sensitive data flows and flags violations of established cryptographic rules. Applying MICRYSCOPE to 9,403 MCP servers, we identified 720 with cryptographic logic, of which 19.7% exhibited misuses. These flaws are concentrated in certain markets (e.g., Smithery Registry with 42% insecure servers), languages (Python at 34% misuse rate), and categories (Developer Tools and Data Science & ML accounting for over 50% of all misuses). Case studies reveal real-world consequences, including leaked API keys, insecure DES/ECB tools, and MD5-based authentication bypasses. Our study establishes the first ecosystem-wide view of cryptographic misuse in MCP and provides both tools and insights to strengthen the security foundations of this rapidly growing protocol.

cs.CR↗

arXiv2025

CCN: Decentralized Cross-Chain Channel Networks Supporting Secure and Privacy-Preserving Multi-Hop Interactions

Cross-chain technology enables interoperability among otherwise isolated blockchains, supporting interactions across heterogeneous networks. Similar to how multi-hop communication became fundamental in the evolution of the Internet, the demand for multi-hop cross-chain interactions is gaining increasing attention. However, this growing demand introduces new security and privacy challenges. On the security side, multi-hop interactions depend on the availability of multiple participating nodes. If any node becomes temporarily offline during execution, the protocol may fail to complete correctly, leading to settlement failure or fund loss. On the privacy side, the need for on-chain transparency to validate intermediate states may unintentionally leak linkable information, compromising the unlinkability of user interactions. In this paper, we propose the Cross-Chain Channel Network (CCN), a decentralized network designed to support secure and privacy-preserving multi-hop cross-chain transactions. Through experimental evaluation, we identify two critical types of offline failures, referred to as active and passive offline cases, which have not been adequately addressed by existing solutions. To mitigate these issues, we introduce R-HTLC, a core protocol within CCN. R-HTLC incorporates an hourglass mechanism and a multi-path refund strategy to ensure settlement correctness even when some nodes go offline during execution. Importantly, CCN addresses not only the correctness under offline conditions but also maintains unlinkability in such adversarial settings. To overcome this, CCN leverages zero-knowledge proofs and off-chain coordination, ensuring that interaction relationships remain indistinguishable even when certain nodes are temporarily offline.

cs.CR↗

arXiv2025

A Measurement Study of Model Context Protocol Ecosystem

The Model Context Protocol (MCP) has been proposed as a unifying standard for connecting large language models (LLMs) with external tools and resources, promising the same role for AI integration that HTTP and USB played for the Web and peripherals. Yet, despite rapid adoption and hype, its trajectory remains uncertain. Are MCP marketplaces truly growing, or merely inflated by placeholders and abandoned prototypes? Are servers secure and privacy-preserving, or do they expose users to systemic risks? And do clients converge on standardized protocols, or remain fragmented across competing designs? In this paper, we present the first large-scale empirical study of the MCP ecosystem. We design and implement MCPCrawler, a systematic measurement framework that collects and normalizes data from six major markets. Over a 14-day campaign, MCPCrawler aggregated 17,630 raw entries, of which 8,401 valid projects (8,060 servers and 341 clients) were analyzed. Our results reveal that more than half of listed projects are invalid or low-value, that servers face structural risks including dependency monocultures and uneven maintenance, and that clients exhibit a transitional phase in protocol and connection patterns. Together, these findings provide the first evidence-based view of the MCP ecosystem, its risks, and its future trajectory.

cs.CY↗

arXiv2025

Assouad and quasi-Assouad dimensions of Moran sets

Moran sets are a non-autonomous generalization of self-similar sets. In this paper, we study the quasi-Assouad and Assouad dimensions of Moran sets in $\mathbb{R}^{d}$. First we provide quasi-Assouad dimension formulae for Moran sets satisfying $c_*>0$. Then, we provide the upper and lower bounds for quasi-Assouad dimension formulae for Moran sets without assuming $c_*>0$. To obtain the exact dimension formulae in this case, we define quasi-normal and normal Moran sets, and provide quasi-Assouad dimension formulae for these sets.

math.DS↗

arXiv2025

What You Code Is What We Prove: Translating BLE App Logic into Formal Models with LLMs for Vulnerability Detection

The application layer of Bluetooth Low Energy (BLE) is a growing source of security vulnerabilities, as developers often neglect to implement critical protections such as encryption, authentication, and freshness. While formal verification offers a principled way to check these properties, the manual effort of constructing formal models makes it impractical for large-scale analysis. This paper introduces a key insight: BLE application security analysis can be reframed as a semantic translation problem, i.e., from real-world code to formal models. We leverage large language models (LLMs) not to directly detect vulnerabilities, but to serve as translators that convert BLE-specific code into process models verifiable by tools like ProVerif. We implement this idea in VerifiaBLE, a system that combines static analysis, prompt-guided LLM translation, and symbolic verification to check three core security features: encryption, randomness, and authentication. Applied to 1,050 Android BLE apps, VerifiaBLE uncovers systemic weaknesses: only 10.2\% of apps implement all three protections, while 53.9\% omit them entirely. Our work demonstrates that using LLMs as structured translators can lower the barrier to formal methods, unlocking scalable verification across security-critical domains.

cs.CR↗

arXiv2025

Dataset Ownership in the Era of Large Language Models

As datasets become critical assets in modern machine learning systems, ensuring robust copyright protection has emerged as an urgent challenge. Traditional legal mechanisms often fail to address the technical complexities of digital data replication and unauthorized use, particularly in opaque or decentralized environments. This survey provides a comprehensive review of technical approaches for dataset copyright protection, systematically categorizing them into three main classes: non-intrusive methods, which detect unauthorized use without modifying data; minimally-intrusive methods, which embed lightweight, reversible changes to enable ownership verification; and maximally-intrusive methods, which apply aggressive data alterations, such as reversible adversarial examples, to enforce usage restrictions. We synthesize key techniques, analyze their strengths and limitations, and highlight open research challenges. This work offers an organized perspective on the current landscape and suggests future directions for developing unified, scalable, and ethically sound solutions to protect datasets in increasingly complex machine learning ecosystems.

cs.CR↗

arXiv2025

LIFT: Automating Symbolic Execution Optimization with Large Language Models for AI Networks

Dynamic Symbolic Execution (DSE) is a key technique in program analysis, widely used in software testing, vulnerability discovery, and formal verification. In distributed AI systems, DSE plays a crucial role in identifying hard-to-detect bugs, especially those arising from complex network communication patterns. However, traditional approaches to symbolic execution are often hindered by scalability issues and inefficiencies, particularly in large-scale systems. This paper introduces LIFT (Large-language-model Integrated Functional-equivalent-IR Transformation), a novel framework that leverages Large Language Models (LLMs) to automate the optimization of Intermediate Representations (IRs) in symbolic execution. LIFT addresses the challenges of symbolic execution by providing a scalable, context-sensitive solution for IR transformation. The framework consists of two phases: IR Analysis and Optimization, where LLMs optimize time-intensive IR blocks, and Symbolic Execution and Validation, which includes benchmarking and semantic verification to ensure correctness and generalizability. Experiments on real-world binaries demonstrated significant performance improvements, including a 53.5\% reduction in execution time for bigtest and a 10.24\% reduction for random, along with reductions in IR statements, PUT instructions, and temporary variables. These results demonstrate that LLMs simplify IRs while maintaining functional correctness, enhancing symbolic execution in distributed AI systems.

cs.CR↗

arXiv2025

We Urgently Need Privilege Management in MCP: A Measurement of API Usage in MCP Ecosystems

The Model Context Protocol (MCP) has emerged as a widely adopted mechanism for connecting large language models to external tools and resources. While MCP promises seamless extensibility and rich integrations, it also introduces a substantially expanded attack surface: any plugin can inherit broad system privileges with minimal isolation or oversight. In this work, we conduct the first large-scale empirical analysis of MCP security risks. We develop an automated static analysis framework and systematically examine 2,562 real-world MCP applications spanning 23 functional categories. Our measurements reveal that network and system resource APIs dominate usage patterns, affecting 1,438 and 1,237 servers respectively, while file and memory resources are less frequent but still significant. We find that Developer Tools and API Development plugins are the most API-intensive, and that less popular plugins often contain disproportionately high-risk operations. Through concrete case studies, we demonstrate how insufficient privilege separation enables privilege escalation, misinformation propagation, and data tampering. Based on these findings, we propose a detailed taxonomy of MCP resource access, quantify security-relevant API usage, and identify open challenges for building safer MCP ecosystems, including dynamic permission models and automated trust assessment.

cs.CR↗

arXiv2025

Can Large Language Models Be Trusted Paper Reviewers? A Feasibility Study

Academic paper review typically requires substantial time, expertise, and human resources. Large Language Models (LLMs) present a promising method for automating the review process due to their extensive training data, broad knowledge base, and relatively low usage cost. This work explores the feasibility of using LLMs for academic paper review by proposing an automated review system. The system integrates Retrieval Augmented Generation (RAG), the AutoGen multi-agent system, and Chain-of-Thought prompting to support tasks such as format checking, standardized evaluation, comment generation, and scoring. Experiments conducted on 290 submissions from the WASA 2024 conference using GPT-4o show that LLM-based review significantly reduces review time (average 2.48 hours) and cost (average \$104.28 USD). However, the similarity between LLM-selected papers and actual accepted papers remains low (average 38.6\%), indicating issues such as hallucination, lack of independent judgment, and retrieval preferences. Therefore, it is recommended to use LLMs as assistive tools to support human reviewers, rather than to replace them.

cs.CY↗

arXiv2025

ClusterUCB: Efficient Gradient-Based Data Selection for Targeted Fine-Tuning of LLMs

Gradient-based data influence approximation has been leveraged to select useful data samples in the supervised fine-tuning of large language models. However, the computation of gradients throughout the fine-tuning process requires too many resources to be feasible in practice. In this paper, we propose an efficient gradient-based data selection framework with clustering and a modified Upper Confidence Bound (UCB) algorithm. Based on the intuition that data samples with similar gradient features will have similar influences, we first perform clustering on the training data pool. Then, we frame the inter-cluster data selection as a constrained computing budget allocation problem and consider it a multi-armed bandit problem. A modified UCB algorithm is leveraged to solve this problem. Specifically, during the iterative sampling process, historical data influence information is recorded to directly estimate the distributions of each cluster, and a cold start is adopted to balance exploration and exploitation. Experimental results on various benchmarks show that our proposed framework, ClusterUCB, can achieve comparable results to the original gradient-based data selection methods while greatly reducing computing consumption.

cs.CL↗

arXiv2025

Self-Error-Instruct: Generalizing from Errors for LLMs Mathematical Reasoning

Although large language models demonstrate strong performance across various domains, they still struggle with numerous bad cases in mathematical reasoning. Previous approaches to learning from errors synthesize training data by solely extrapolating from isolated bad cases, thereby failing to generalize the extensive patterns inherent within these cases. This paper presents Self-Error-Instruct (SEI), a framework that addresses these model weaknesses and synthesizes more generalized targeted training data. Specifically, we explore a target model on two mathematical datasets, GSM8K and MATH, to pinpoint bad cases. Then, we generate error keyphrases for these cases based on the instructor model's (GPT-4o) analysis and identify error types by clustering these keyphrases. Next, we sample a few bad cases during each generation for each identified error type and input them into the instructor model, which synthesizes additional training data using a self-instruct approach. This new data is refined through a one-shot learning process to ensure that only the most effective examples are kept. Finally, we use these curated data to fine-tune the target model, iteratively repeating the process to enhance performance. We apply our framework to various models and observe improvements in their reasoning abilities across both in-domain and out-of-domain mathematics datasets. These results demonstrate the effectiveness of self-error instruction in improving LLMs' mathematical reasoning through error generalization.

cs.CL↗

arXiv2025

Say What You Mean: Natural Language Access Control with Large Language Models for Internet of Things

Access control in the Internet of Things (IoT) is becoming increasingly complex, as policies must account for dynamic and contextual factors such as time, location, user behavior, and environmental conditions. However, existing platforms either offer only coarse-grained controls or rely on rigid rule matching, making them ill-suited for semantically rich or ambiguous access scenarios. Moreover, the policy authoring process remains fragmented: domain experts describe requirements in natural language, but developers must manually translate them into code, introducing semantic gaps and potential misconfiguration. In this work, we present LACE, the Language-based Access Control Engine, a hybrid framework that leverages large language models (LLMs) to bridge the gap between human intent and machine-enforceable logic. LACE combines prompt-guided policy generation, retrieval-augmented reasoning, and formal validation to support expressive, interpretable, and verifiable access control. It enables users to specify policies in natural language, automatically translates them into structured rules, validates semantic correctness, and makes access decisions using a hybrid LLM-rule-based engine. We evaluate LACE in smart home environments through extensive experiments. LACE achieves 100% correctness in verified policy generation and up to 88% decision accuracy with 0.79 F1-score using DeepSeek-V3, outperforming baselines such as GPT-3.5 and Gemini. The system also demonstrates strong scalability under increasing policy volume and request concurrency. Our results highlight LACE's potential to enable secure, flexible, and user-friendly access control across real-world IoT platforms.

cs.CL↗

arXiv2025

Everything You Wanted to Know About LLM-based Vulnerability Detection But Were Afraid to Ask

Large Language Models are a promising tool for automated vulnerability detection, thanks to their success in code generation and repair. However, despite widespread adoption, a critical question remains: Are LLMs truly effective at detecting real-world vulnerabilities? Current evaluations, which often assess models on isolated functions or files, ignore the broader execution and data-flow context essential for understanding vulnerabilities. This oversight leads to two types of misleading outcomes: incorrect conclusions and flawed rationales, collectively undermining the reliability of prior assessments. Therefore, in this paper, we challenge three widely held community beliefs: that LLMs are (i) unreliable, (ii) insensitive to code patches, and (iii) performance-plateaued across model scales. We argue that these beliefs are artifacts of context-deprived evaluations. To address this, we propose CORRECT (Context-Rich Reasoning Evaluation of Code with Trust), a new evaluation framework that systematically incorporates contextual information into LLM-based vulnerability detection. We construct a context-rich dataset of 2,000 vulnerable-patched program pairs spanning 99 CWEs and evaluate 13 LLMs across four model families. Our framework elicits both binary predictions and natural-language rationales, which are further validated using LLM-as-a-judge techniques. Our findings overturn existing misconceptions. When provided with sufficient context, SOTA LLMs achieve significantly improved performance (e.g., 0.7 F1-score on key CWEs), with 0.8 precision. We show that most false positives stem from reasoning errors rather than misclassification, and that while model and test-time scaling improve performance, they introduce diminishing returns and trade-offs in recall. Finally, we uncover new flaws in current LLM-based detection systems, such as limited generalization and overthinking biases.

cs.CR↗

arXiv2025

Pangu Ultra: Pushing the Limits of Dense Large Language Models on Ascend NPUs

We present Pangu Ultra, a Large Language Model (LLM) with 135 billion parameters and dense Transformer modules trained on Ascend Neural Processing Units (NPUs). Although the field of LLM has been witnessing unprecedented advances in pushing the scale and capability of LLM in recent years, training such a large-scale model still involves significant optimization and system challenges. To stabilize the training process, we propose depth-scaled sandwich normalization, which effectively eliminates loss spikes during the training process of deep models. We pre-train our model on 13.2 trillion diverse and high-quality tokens and further enhance its reasoning capabilities during post-training. To perform such large-scale training efficiently, we utilize 8,192 Ascend NPUs with a series of system optimizations. Evaluations on multiple diverse benchmarks indicate that Pangu Ultra significantly advances the state-of-the-art capabilities of dense LLMs such as Llama 405B and Mistral Large 2, and even achieves competitive results with DeepSeek-R1, whose sparse model structure contains much more parameters. Our exploration demonstrates that Ascend NPUs are capable of efficiently and effectively training dense models with more than 100 billion parameters. Our model and system will be available for our commercial customers.

cs.CL↗

arXiv2025

Asynchronous BFT Consensus Made Wireless

Asynchronous Byzantine fault-tolerant (BFT) consensus protocols, known for their robustness in unpredictable environments without relying on timing assumptions, are becoming increasingly vital for wireless applications. While these protocols have proven effective in wired networks, their adaptation to wireless environments presents significant challenges. Asynchronous BFT consensus, characterized by its N parallel consensus components (e.g., asynchronous Byzantine agreement, reliable broadcast), suffers from high message complexity, leading to network congestion and inefficiency, especially in resource-constrained wireless networks. Asynchronous Byzantine agreement (ABA) protocols, a foundational component of asynchronous BFT, require careful balancing of message complexity and cryptographic overhead to achieve efficient implementation in wireless settings. Additionally, the absence of dedicated testbeds for asynchronous wireless BFT consensus protocols hinders development and performance evaluation. To address these challenges, we propose a consensus batching protocol (ConsensusBatcher), which supports both vertical and horizontal batching of multiple parallel consensus components. We leverage ConsensusBatcher to adapt three asynchronous BFT consensus protocols (HoneyBadgerBFT, BEAT, and Dumbo) from wired networks to resource-constrained wireless networks. To evaluate the performance of ConsensusBatcher-enabled consensus protocols in wireless environments, we develop and open-source a testbed for deployment and performance assessment of these protocols. Using this testbed, we demonstrate that ConsensusBatcher-based consensus reduces latency by 48% to 59% and increases throughput by 48% to 62% compared to baseline consensus protocols.

cs.DC↗