Skip to content
View yuxuan-z19's full-sized avatar
  • Tsinghua University
  • Beijing, China

Highlights

  • Pro

Block or report yuxuan-z19

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Maximum 250 characters. Please don't include any personal information such as legal names or email addresses. Markdown supported. This note will be visible to only you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
yuxuan-z19/README.md

Ex-PERF, LLM codesign newbie

"No further hesitation / On those unanswered questions / So now, I'll make a dream unchained"

About Me

Benchmarked neuromorphic systems for 3 years; now wrestling with LLM performance on heterogenenous clusters:

Research Interests 🔬 Hobbies 🧩
- Making LLMs faster across heterogeneous clusters
- Smarter AI compiler and kernel generation
- Fair benchmarking with Double-ML + AutoML
- Honkai: Star Rail & BanG Dream!
- Solo-leveling from E-rank in LLM codesign

For more information, see resume (CN)

Education

  • {2023.06-Now} M.Eng. in Computer Science and Technology (Major in Computer Systems Organization), supervised by Prof. Youhui Zhang with CRAFT Lab, Dept. CST, Tsinghua University, China.
  • {2019.08-2023.06} B.Eng. in Computer Science and Technology, Dept. CST, Tsinghua University, China.

Experiences

Interns

  • {2025.06-2025.09} LLM Research Intern, Model R&D Division, AI Cloud Group (ACG), Baidu AI Cloud
    • Focused on CUDA kernel generation driven by self-evolving coding agents
  • {2022.06-2022.09} RTL Design Intern, Heterogeneous Computing Division, Kuaishou Technology
    • Engaged in RTL prototyping and design of custom AI accelerators developed in-house
    • 220831: Received the "Best Intern Award" (Top 3 Recipients)

Services

  • {2025.04-Now} Huawei Campus Ambassador
  • {2025-2026 Fall} TA for "Big Data and Machine Intelligence" / 《大数据与机器智能》 (01510243), iCenter.
  • {2023-2024 Spring, 2024-2025 Fall} General Office Assistant at Humanities & Social Sciences Library.
  • {2023-2024 Summer} TA for "Innovation Practice of Technology Products"/《科技产品创新实践》 (31510253), iCenter.
  • {2023-2024 Fall} TA for "Introduction to Computer Systems"/《计算机系统概论》 (30240593), Dept. CST.
Cocurricular

Works

Publications

arXiv's

Projects

  • diffonnx - A powerful yet playful tool to compare and analyze ONNX models – whether you're hunting for hidden changes or debugging mysterious outputs

  • codegen-eval - A lightweight, high-concurrency evaluation harness for batch code generation with large language models (LLMs).

Pinned Loading

  1. diffonnx diffonnx Public

    Diff your ONNXs

    Python 2