๐ฟ Lyumanshan Ye (ๅถ็ปฟๆปกๅฑฑ)
Hi, ๐๐ป I'm Lyumanshan Ye, a master's student at Shanghai Jiao Tong University 
. Currently, I am a research intern at the
GAIR Lab 
.
I am fortunate to be advised by
Prof. Pengfei Liu, a mentor I deeply respect and am sincerely grateful to. His full support has been instrumental in shaping my exploration at the intersection of NLP and HCI. Here are some core problems I'm thinking about:
- ๐ชป Human-AI Collaborative & Adaptive Systems: How to optimize task delegation to AI while enhancing the space for human decision-making?
- ๐ Human-Machine Autonomy: How can AI systems integrate with human workflows while preserving human autonomy and control?
- ๐ Scaffold Data to Improve Foundation Reasoning Ability: How to use data from human-AI scaffold data to train agents to learn complex rewards?
My ultimate research goal is to develop Human-AI communication layers that enhance users' ability to articulate their context, maximize collaborative performance, and enable AI systems to learn from real-world task rewards. In my view, AI should serve as a deep partner to humans, augmenting their capabilities rather than replacing them. As a researcher with an interdisciplinary background, my MSc major is Mechanical Engineering from SJTU's School of Design (advised by Prof. Danni Chang). I have collaborated with Prof. Xin Yi (Tsinghua University) on LLM privacy and Dr. Qiao Jin (Carnegie Mellon University) on AIED.
A few interesting facts about me: My name has been changed three times, as my family wished it to convey the freshness and vitality of spring. In Chinese, it means "green leaves covering the mountains," a fitting image of my hometown, Fujian โ a region known for its verdant hills.
Research
The papers below are sorted by the time of research. My research journey has been anything but linear. If these research questions resonate with you, I'd love to hear from you. Click here to reach out :D
News / Travel
Apr 2025 ๐ ๐ฏ๐ต CHI 2025 Yokohama, Japan ๐ธ ๐ผ