I am a research intern at the Penn HCI Group, advised by Dr. Andrew Head. My research direction is autonomous research for human-computer interaction. I'm currently developing an automatic evaluation tool for programming language usability with simulated users. In this project, I developed a cognitive-aware coder agent to simulate programmers' behavior, an interactive UI to facilitate sensemaking processes for programming traces, a dataset incorporates real-world human programming traces for benchmarking, and a crowdsourcing platform for programming trace collections.
I am a research intern at the Northwestern MU Collective Group, advised by Dr. Jessica Hullman and Ziyang Guo, PhD Candidate. My research direction is human-AI collaborative decision making in scientific research. I'm currently developing a benchmark and a signal modeling framework to evaluate the human-AI collaboration values through statistical modeling. In the project, I developed an end-to-end evaluation pipeline for benchmarking, a set of LLM-driven inference pipelines (agentic, claim-based, cluster-based), a large-scale data collection and processing pipeline, and a statistical modeling pipeline.
I am responsible for the technical development of an agentic system to simulate AI scientists. I am working on a multi-agentic pipeline to accelerate scientific experiment execution. I conduct computational experiments to benchmark our proposed systems and design use cases to demonstrate the usability and effectiveness of our system.
I'm a Research Team Lead at the Visual Intelligence Lab, UMNTC, where I am advised by Prof. Qianwen Wang. My primary research areas are human-computer interactions, visualization, and human-centered AI systems. I am also actively working on sub-domains such as cognitive/behavioral modeling, gamification, and educational technologies.
As a Research Team Lead, I have successfully completed three Research and development projects and produced two full papers, one short paper, and one workshop presentation, all as first authors. My major works included an interaction tool for GNNs, a sensemaking game for LLMAS, a Python package for GNN visualization, and a cognitive modeling toolkit for scientific writing.
Besides my scientific research work, my other responsibilities included interviewing/mentoring new members, project management, and planning.
I was responsible for the technical development of an internal crowdsourcing system for asteroid classification, which helps the training process of our machine learning algorithms. I also developed several internal tool utilities for the TURBO Telescopes, including an image processing pipeline, an internal watchdog system, and an image storage module.
I was responsible for the technical development of a motion capture interface in VR on the Oculus platform. I utilized tools such as Unity 3D, Blender, and OpenXR Toolkit. The project included a humanoid model integrated with inverse kinematic and a motion capture system.