Skip to content
View JianhaoYu-AI's full-sized avatar
🎯
Focusing
🎯
Focusing
0 followers · 4 following

Block or report JianhaoYu-AI

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.

Content in all repositories owned by your account will be closed.
Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

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

Report abuse
JianhaoYu-AI/README.md

Waving hand Hi, I'm Jianhao Yu.

Studying agent harnesses and evaluation for reliable AI agents.

🎓 M.S. Student in AI · Beijing Jiaotong University 📍 Based in Beijing, China

💼 Research & Experience

Research interests
Agent Harness · Agent Evaluation · Reliable AI Agents
Exploring agentic RL and LLM post-training, with a background in LLM reasoning and multimodal applications.

Shanghai AI Laboratory
LLM & Agent Algorithms Intern · Dec 2025–Apr 2026
Scientific Agent training, evaluation, and GPU post-training experiments.

Project stack
Python · PyTorch · Kotlin · Jetpack Compose · Node.js

💻 Projects & Open Source

NetScope
Android network diagnostics with live probes, evidence snapshots, budgeted tool calls, citation checks, and before/after measurements.
Kotlin/Compose + Node.js; JUnit and gateway tests.

SkillOpt — Fork of microsoft/SkillOpt; studying validation-gated agent skill optimization.

Driver Fatigue Detection — Earlier computer-vision project with YOLO11 Web and YOLOv12 desktop implementations, trained weights, and historical training logs.

🏅 Awards, Experience & Research →

GitHub · Academic homepage · Email

Pinned Loading

  1. NetScope NetScope Public

    Kotlin 1

  2. driver-fatigue-detection-system driver-fatigue-detection-system Public

    YOLO11 Flask web and YOLOv12 PyQt5 desktop driver fatigue detection, with trained weights and historical training records.

    Python

  3. SkillOpt SkillOpt Public

    Forked from microsoft/SkillOpt

    SkillOpt is a text-space optimizer that trains reusable natural-language skills for frozen LLM agents through trajectory-driven edits, validation-gated updates, and deployable best_skill.md artifacts.

    Python

  4. JianhaoYu-AI.github.io JianhaoYu-AI.github.io Public

    Personal academic homepage of Yu Jianhao.

    HTML