Skip to content
View EachSheep's full-sized avatar
🎯
Focusing
🎯
Focusing

Highlights

  • Pro

Block or report EachSheep

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 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
eachsheep/README.md

Hi thereπŸ‘‹

GitHub RSS

Short Bio

Important

I am currently actively seeking job opportunities related to LLM agents, software engineering, and real-world automation. Welcome to email me at hyshen@stu.pku.edu.cn if you are interested.

  • πŸŽ“ I am SHEN Haiyang, a Ph.D. candidate at the School of Computer Science, Peking University, advised by Assistant Professor Yun Ma.

  • πŸ… I received my B.Sc. degree from the School of Computer Science at Northwestern Polytechnical University in June 2022.

  • 🌱 I work on LLM agents, software engineering, and data synthesis for real-world automation. My current research takes a data-centric view of agent development: the bottleneck is not only model capability, but also how we discover, define, and scale the right tasks from real software systems, user trajectories, and executable environments. Since late 2024, I have focused on agent data synthesis, including graph/search data synthesis for deep research and agentic retrieval, and coding agent benchmark/data synthesis through Monthly-SWEBench, DeepTerminalBench, and EvoCodeBench.

  • ♻️ My long-term vision is to build a fully automated self-improving loop for agents: systems that discover real problems from software systems and user trajectories, define them as measurable tasks, synthesize benchmark and training data, run experiments, and use feedback to improve the next generation of agents. Inspired by autoresearch, I hope to move this loop toward fully automated agent self-evolution and continual evolution.

  • πŸ“š My research group is affiliated with the Data Space Technology and Systems Research Center, led by Academician Hong Mei and Professor Gang Huang, with faculty members including Xuanzhe Liu, Xin Jin, and Yun Ma.

Recent Experience

  • 2025.12 -- 2026.07, Intern at UniPat AI. Worked on high-quality coding-agent benchmarks and data for Terminal-X, including Deep-TerminalBench, EvoCode-Bench, Roadmap-Bench, and Monthly-SWEBench. Beyond human-in-the-loop benchmark construction, I advanced fully automated, large-scale high-quality data construction pipelines that have already created substantial economic value.

πŸ“« Correspondence

🌟 Others

  • πŸ–‹ My introduction at here.

πŸ““ Gists I wrote

⭐ Recent Stars

Pinned Loading

  1. UniPat-AI/EvoCodeBench UniPat-AI/EvoCodeBench Public

    Python 13

  2. UniPat-AI/Terminal-X UniPat-AI/Terminal-X Public

    Official repository of blog: Terminal-X: Evaluate Coding Agents across Depth, Iteration, and Evolution in Terminal Environments

    Rust

  3. gta0804/AlphaPROBE gta0804/AlphaPROBE Public

    Official implementation of AlphaPROBE: Alpha mining with principled evolution and on-graph biased evolution

    Python 86 27

  4. Alibaba-NLP/DeepResearch Alibaba-NLP/DeepResearch Public

    Tongyi Deep Research, the Leading Open-source Deep Research Agent

    Python 19.7k 1.5k

  5. gta0804/MASS gta0804/MASS Public

    Official implementation of MASS: Multi-Agent Simulation Scaling for Portfolio Construction

    Python 178 25

  6. ShortcutsBench ShortcutsBench Public

    ShortcutsBench: A Large-Scale Real-World Benchmark for API-Based Agents

    Python 111 6