Work Experience

AI Algorithm Department Supervisor: Asst. Prof. Yongjian Zhao

• Proposed DCT-based frequency domain visual servoing algorithm for robotic PAT system
• Achieved 98.7% image similarity with 18.84% speed improvement over traditional DVS methods

Advanced Technology Department Supervisor: Prof. Zuotai Zhang

• Designed SciBERT-CRF weak supervision model for materials science literature processing
• Developed end-to-end AI pipeline with four-stage progressive optimization
• Achieved F1-score of 0.791, precision improved from 0.359 to 0.856

Infectious Disease Prevention Institute Supervisors: Director Yanpeng Cheng, Assoc. Prof. Zhenghui Feng

• Developed dual-dimension clustering analysis framework for COVID-19 transmission heterogeneity
• Quantified super-spreading phenomenon: 13.7% transmission chains caused 80% of total infections

Frontier Intelligence Department

Key Responsibilities:

• Model Fine-tuning: Fine-tune LLM/VLM/VLA foundation models to explore planning and decision-making capabilities for instrument tasks, forming verifiable technical roadmaps

• Data & Evaluation System: Build training datasets, offline evaluation, and regression mechanisms to drive quantifiable iteration

• Fine-tuning & Alignment: Hands-on experience with SFT and parameter-efficient fine-tuning to improve task success rate and stability

• RL Practice: Design reward and feedback loops to advance policy optimization effectiveness in real-world tasks

• Engineering Collaboration: Work with engineering teams to ensure models are deployable, monitorable, and iteratively improvable in production systems