Yaofeng Su

GPA: 3.83/4.00 (High academic standing within major). Recipient of Fudan University Outstanding Student Award (with scholarship). Achieved 4th place in ASIC National Key Laboratory Hackathon. Recognized with numerous other significant awards and scholarships.

Intern at MagicLab, School of Engineering and Applied Technology, leading an autonomous driving risk assessment research project under the guidance of Professor Wenchao Ding.

Scheduled to conduct research on vision-language-action models under Professor Huaxiu Yao, focusing on Supervised Fine-Tuning (SFT) and Reinforcement Learning (RL) approaches. Co-authored a paper submitted to a top-tier AI conference.

Status: Submitted to a top-tier AI conference (details pending double-blind review).

• Trained Vision-Language-Action (VLA) models using novel Reinforcement Learning (RL) methodologies.
• Designed and executed experiments, including modifying tasks within simulated environments to evaluate model capabilities.
• Managed dataset collection, processing, and evaluation for VLA model training and validation.

Status: Draft phase; focused on gaining research experience and training.

• Engineered risk field representations using advanced spatiotemporal modeling techniques.
• Developed and implemented realistic traffic scene generation leveraging diffusion models combined with gradient optimization.
• Designed and actualized a lightweight neural network for efficient risk prediction.