Institution profile

Washington University in St. Louis

Academic institutionnorthamerica · us
Official website
Research library531linked papers
Opportunities0open roles
Selected work

Representative Papers

Measuring Defi Risk

Sep 07, 2026Social Science Research Network

本文针对DeFi借贷风险问题，通过构建一个仅需总存款和借款数据的框架来评估系统整体风险，为投资者提供预警。

11 citations2 influentialRead paper

Performance Analysis of OpenVPN on a Consumer Grade Router

Apr 27, 2025

This study addresses performance bottlenecks of OpenVPN on resource-constrained Linksys WRT54GL routers. It systematically investigates the impact of encryption algorithms (AES-128, Blowfish, 3DES) and transport protocols (TCP/UDP) on throughput and round-trip time (RTT). Methodologically, it employs a $2^{5-1}$ fractional factorial design—the first such application in this context—to quantify main and interaction effects of five key factors. Empirical evaluation is conducted on DD-WRT firmware with OpenSSL-based OpenVPN configurations. Results demonstrate that cryptographic overhead is the dominant throughput bottleneck, with AES-128 achieving optimal trade-offs between security and efficiency; meanwhile, transport protocol selection governs RTT behavior, with UDP yielding significantly lower latency than TCP. This work establishes a reproducible experimental framework and provides empirical guidance for lightweight VPN deployment on embedded systems.

8 citationsRead paper

CoreGuard: Safeguarding Foundational Capabilities of LLMs Against Model Stealing in Edge Deployment

Oct 16, 2024arXiv.org

Deploying proprietary large language models (LLMs) on edge devices faces fundamental capability leakage risks: adversaries can bypass weight-protection mechanisms via advanced attacks such as fine-tuning, while existing trusted execution environment (TEE) solutions incur prohibitive communication and computational overhead, rendering them impractical for edge deployment. This paper proposes CoreGuard—the first lightweight, propagatable TEE-based authorization mechanism specifically designed to protect LLMs’ foundational capabilities (rather than task-specific parameters). CoreGuard achieves core capability isolation with minimal overhead through three key innovations: capability abstraction, lightweight authorization protocols, and in-TEE propagation control. Experimental evaluation demonstrates that CoreGuard delivers black-box–equivalent security, incurs less than 0.5% inference latency overhead, reduces TEE–CPU communication volume by 92%, and enables real-time edge deployment—thereby overcoming the critical bottleneck hindering TEE adoption for edge-hosted LLMs.

6 citationsRead paper

Towards Global Optimality in Cooperative MARL with the Transformation And Distillation Framework

Jul 12, 2022

In decentralized execution for cooperative multi-agent reinforcement learning (MARL), mainstream decentralized policy gradient methods suffer from inherent suboptimality, preventing convergence to globally optimal policies. Method: We propose the Transformation-and-Distillation (TAD) framework, which equivalently reformulates a cooperative multi-agent MDP into a sequential single-agent MDP and employs policy distillation to recover decentralized execution. Contribution/Results: We theoretically prove that TAD guarantees learning of globally optimal policies in finite MDPs. Instantiating TAD with PPO, we develop TAD-PPO—incorporating MDP structural transformation, two-stage training, and value decomposition analysis. Empirical evaluation across diverse cooperative benchmarks demonstrates that TAD-PPO significantly outperforms state-of-the-art methods, achieving both theoretical global optimality guarantees and strong generalization capability.

5 citationsRead paper

Benchmark^2: Systematic Evaluation of LLM Benchmarks

Jan 07, 2026arXiv.org

This work addresses the proliferation of large language model (LLM) evaluation benchmarks, which has outpaced systematic assessment of their intrinsic quality. To this end, we propose Benchmark², a novel framework that establishes the first quantitative methodology for evaluating the reliability and validity of LLM benchmarks through three complementary metrics: cross-benchmark ranking consistency, discriminability score, and capability alignment bias. Empirical evaluation across 15 benchmarks and 11 LLMs demonstrates that Benchmark² not only reveals substantial quality disparities among existing benchmarks but also enables the construction of streamlined test sets that maintain high evaluative performance while significantly reducing assessment scale.

3 citationsRead paper
Recent publications

Latest Papers