Ph.D. Student · CUHK CSE

Building efficient, reliable AI systems.

I am Yuedong Zhong, a Ph.D. student in Computer Science and Engineering at The Chinese University of Hong Kong. I am advised by Prof. Michael R. Lyu. My work sits at the intersection of AI systems, high-performance computing, AI agents, and quantitative research.

Before starting my Ph.D., I studied at Sun Yat-sen University and completed research assistant appointments at CUHK in 2025 and 2026. I am also interested in crypto quantitative research.

AI Systems High-Performance Computing AI Agents Reliable LLM Systems Crypto Quantitative Research
View publications Email me
04Selected papers
02Research institutions
04Core research directions

Systems-minded research, from low-bit LLM kernels to reliable AI agents and multi-agent diagnosis.

Background

Experience & education

A path shaped by systems research, engineering practice, and quantitative thinking.

Experience

  1. 2026–Present

    First-year Ph.D. Student

    Department of Computer Science and Engineering, CUHK

  2. 2026.03–07

    Research Assistant

    The Chinese University of Hong Kong

    LLM system bug detection and empirical failure analysis.

  3. 2025.03–09

    Research Assistant

    The Chinese University of Hong Kong

    Ops agents for cloud operations with retrieval-augmented generation.

  4. 2024.05–11

    AI Infrastructure Research Intern

    Tencent Games

  5. 2023.07–09

    C++ Developer Intern

    Wizard Investment

  6. 2022.07–11

    Backend Developer Intern

    Tencent Cloud Games

Education

The Chinese University of Hong Kong

Ph.D. in Computer Science and Engineering

Advisor: Prof. Michael R. Lyu

Hong Kong SAR · 2026–Present

Sun Yat-sen University

Bachelor's degree, School of Software Engineering

Guangzhou, China · 2020–2024
Good systems work turns an ambitious idea into something measurable, dependable, and fast.

Research

Selected publications

Work across dependable AI services, agentic systems, retrieval-augmented systems, and efficient LLM inference.

01
DSN ’26Research Track

LLMGuard: Multi-Agent Fault Diagnosis for Reliable Language-Model-as-a-Service

Yuedong Zhong, Guangba Yu, Yujie Huang, Qunchao Fu, Rui Ren, Cong Feng, Yongqiang Yang, Michael R. Lyu

A multi-agent approach to diagnosing faults in Language-Model-as-a-Service systems, designed to improve reliability in production AI services.

DSN 2026 DOI
2026
02
arXiv ’26LLM Reliability

Why Does the LLM Stop Computing: An Empirical Study of User-Reported Failures in Open-Source LLMs

Guangba Yu, Zirui Wang, Yujie Huang, Renyi Zhong, Yuedong Zhong, Yilun Wang, Michael R. Lyu

A large-scale empirical study of 705 real-world failures across the open-source DeepSeek, Llama, and Qwen ecosystems, revealing systemic reliability barriers in user-managed LLM deployment stacks.

arXiv PDF
2026
03
ASE ’25Distinguished Paper Award

iKnow: an Intent-Guided Chatbot for Cloud Operations with Retrieval-Augmented Generation

Junjie Huang, Yuedong Zhong, Guangba Yu, Zhihan Jiang, Minzhi Yan, Wenfei Luan, Tianyu Yang, Rui Ren, Michael R. Lyu

An intent-guided RAG chatbot for cloud operations that improves answer accuracy through intent detection, query rewriting, and missing-knowledge detection.

Paper Code DOI
2025
04
ICLR ’25Efficient LLMs

STBLLM: Breaking the 1-Bit Barrier with Structured Binary LLMs

Peijie Dong, Lujun Li, Yuedong Zhong, Dayou Du, Ruibo Fan, Yuhan Chen, Zhenheng Tang, Qiang Wang, Wei Xue, Yike Guo, Xiaowen Chu

Structured N:M binarization pushes LLM compression below one bit per weight, paired with layer-wise allocation and a specialized CUDA kernel.

Paper arXiv Code
2025