💻 Internships

• 2025.03 - 2026.09, Research Assistant, XYZ AI Lab, Agentic RL Group, Beijing. (Mentors: Jiang Bian and Lei Song)
  • Worked on multi-step search trajectory distillation and curation for SFT and GRPO experiments, including tool-call consistency, evidence checks, deduplication, and trace mixing.
  • Developed agent harness and replay support for context compaction, tool-error recovery, and export of training examples aligned with the agent-visible context.
• 2025.10 - 2026.03, Research Assistant, Microsoft Research Asia, Machine Learning Group, Beijing. (Mentors: Jiang Bian and Lei Song)
• 2024.03 - 2024.06, Research Assistant, Microsoft Research Asia, Social Computing Group, Beijing. Worked on LLM safety with Xing Xie; received the Stars of Tomorrow award. Certificate
• 2023.10, Undergraduate Researcher, CRIPAC-DIG, Beijing. Worked on large vision-language model knowledge editing.