# Jinyu Xiang

> Algorithm Researcher at Xiaomi Mimo-Core · AI Agent Researcher · Musician. Research on LLM-based Agents — particularly Agent Self-Evolution and Foundation Agents.

This is the personal homepage of Jinyu Xiang. Key facts:
- Current: Algorithm Researcher at Xiaomi Mimo-Core (LLMs for reasoning, coding, and agents)
- Contact: xiangjy2001@gmail.com
- GitHub: https://github.com/XiangJinyu
- Google Scholar: https://scholar.google.com/citations?user=NJv62qoAAAAJ

## Machine-friendly endpoints

- [Plain-text resume](https://xiangjinyu.github.io/resume.txt): full CV as plain text — prefer this over parsing HTML
- [RSS feed](https://xiangjinyu.github.io/rss.xml): all blog posts with abstracts
- [Sitemap](https://xiangjinyu.github.io/sitemap-index.xml): all pages

## Publications

- [AFlow: Automating Agentic Workflow Generation](https://openreview.net/forum?id=z5uVAKwmjf): ICLR 2025 Oral
- [Self-Supervised Prompt Optimization](https://arxiv.org/abs/2502.06855): EMNLP 2025
- [OpenManus: An Open-Source Framework for Building General AI Agents](https://github.com/FoundationAgents/OpenManus): 50K+ Stars
- [Advances And Challenges In Foundation Agents](https://arxiv.org/abs/2504.01990): Arxiv 2025
- [ReCode: Unify Plan and Action for Universal Granularity Control](https://arxiv.org/abs/2510.23564): Arxiv 2025
- [A Survey of Self-Evolving Agents: On Path to Artificial Super Intelligence](https://arxiv.org/abs/2507.21046): Arxiv 2025
- [MedAgentsBench: Benchmarking Thinking Models and Agent Frameworks for Complex Medical Reasoning](https://arxiv.org/abs/2503.07459): Arxiv 2025
- [AOrchestra: Automating Sub-Agent Creation for Agentic Orchestration](https://arxiv.org/abs/2602.03786): Arxiv 2026
- [AutoEnv: Automated Environments for Measuring Cross-Environment Agent Learning](https://arxiv.org/abs/2511.19304): Arxiv 2025

## Blog posts

- [2025 年终杂谈：关于Agent的思考](https://xiangjinyu.github.io/blog/2025-ai-thoughts-zh): 围绕 Agent Learning/Evolving、AI 产品的方法论、行业竞争与长期壁垒，以及人如何在不确定中重构“奖励”，展开的一篇年终杂谈。
- [End-of-2025 Reflections: Thoughts on Agents](https://xiangjinyu.github.io/blog/2025-ai-thoughts): An end-of-year reflection centered on Agent Learning/Evolving, methodologies for AI products, industry competition and long-term moats, and how people rewire “rewards” amid uncertainty.

## Projects

- [Open Research](https://xiangjinyu.github.io/open-research): Author. An AI-powered research assistant that autonomously conducts end-to-end research — from literature review to experimentation to paper writing. Generated 7 full papers across AI, astrophysics, and philosophy.
- [OpenManus](https://github.com/FoundationAgents/OpenManus): Github 50K+ stars, Founder. Proposed and assembled a team to recreate Manus functionality, completing the project in just three hours, which garnered 35k stars within its first week.
- [Zhihu AI Social Experiment](https://www.zhihu.com/people/-qw-28): Author. Conducted a social experiment by creating a fully autonomous AI account on Zhihu that posted content for a month without human intervention or detection.
- [2024 Alibaba Global Mathematics Competition AI Challenge](https://github.com/XiangJinyu/Math-Multi-Agent): Top 1% (2/563), Author. Developed the MathAI multi-agent system to tackle Olympiad level math problems, reaching the finals and placing second.

## Notes for agents

- All pages are static HTML; no JavaScript is required to read content.
- The homepage (/) contains: about, news, publications, experience, projects, talks, and recent posts, in that order.
- Blog posts live under /blog/<slug>. A Chinese translation, when available, uses the -zh suffix.