靡不有初，鲜克有终。
- 🔬 Open-source developer and researcher focused on Agent RL, LLM post-training, and AI agents.
- 🌟 Developer and maintainer of open-source projects with 100,000+ GitHub stars in total.
- 🧪 Researcher at Emotion Machine Lab, also working on growth and developer/community operations.
- 📫 Reach me at kmno4-song@foxmail.com.
- Agent RL — reinforcement learning for reasoning, tool use, and multi-turn agents.
- LLM Post-training — alignment, feedback mechanisms, and efficient post-training methods.
- AI Agents — agentic systems, research agents, and evaluation.
- Happy-LLM — A from-scratch guide to LLM fundamentals and implementation.
- self-llm — A practical guide to deploying and fine-tuning open-source LLMs.
- tiny-universe — Hands-on implementations of RAG, agents, evaluation, and other LLM systems.
- llm-agent-rl-lab — Reproductions and studies of RL algorithms for LLM agents.
- paper-insight — An AI-powered platform for paper analysis and research workflows.
- huanhuan-chat — A ChatGLM-based character chatbot fine-tuned on dialogue from Empresses in the Palace.
- AMchat — An LLM-powered assistant for advanced mathematics.
- d2l-ai-solutions-manual — Community solutions for Dive into Deep Learning exercises.
- prompt-engineering-for-developers — Chinese learning materials for Andrew Ng's prompt engineering courses.
- InternLM Tutorial — Hands-on tutorials and demos for the InternLM ecosystem.
- Tianji — An open-source system for building socially intelligent AI agents.
Present
- Conduct research on Agent RL, LLM post-training, and agentic systems.
- Work across developer growth, community operations, and the open-source ecosystem.
- Researched multimodal model behavior and safety with Prof. Chi Zhang; first-authored Hard to Read, Easy to Jailbreak at Findings of ACL 2026.
Jun 2024 – Aug 2024
- Studied LLM applications in urban governance and co-authored What can LLM tell us about cities?.
- Hard to Read, Easy to Jailbreak: How Visual Degradation Bypasses MLLM Safety Alignment — Findings of ACL 2026. (First author)
- What can LLM tell us about cities? — arXiv, 2024.
- A Raster Voronoi Diagram Generating Algorithm Using Edge Attribution and Bilateral Scanning — Geomatics and Information Science of Wuhan University, 2024.