About
Xin Huang is an MSc student in Computer Science at the Technical University of Denmark, a thesis student at CSPaper, and a research intern at Microsoft. Across these roles, Xin works on LLM agents, with a focus on research verification, recursive self-improvement, and auto-research. Xin is broadly interested in how agents can become more capable while staying verifiable and trustworthy.
Research interests
- LLM AgentsDesigning and evaluating reliable agentic AI systems.
- Recursive Self-ImprovementStudying how AI systems can improve themselves reliably and verifiably.
- Research VerificationUsing AI systems to verify the soundness of scientific research.
- Auto-ResearchInvestigating open problems in end-to-end automated research systems.
Contact
- xinhuangcs [at] hotmail [dot] com
- Website · xinhuang.me
- GitHub · github.com