SetupX
Experience-driven repository setup with reusable XPU knowledge, speculative execution and rollback, and a prosecutor–judge verification protocol.
Selected papers and research software.
Experience-driven repository setup with reusable XPU knowledge, speculative execution and rollback, and a prosecutor–judge verification protocol.
A time-aware evaluation infrastructure for forecasting agents, with a planner–solver–aggregator framework for gathering evidence and producing forecasts.
Evaluating agents on realistic workspaces that require reasoning across large-scale file dependencies, with task-specific dependency graphs and rubric-based assessment.
A survey of LLM-powered data analysis across structured, semi-structured, unstructured, and heterogeneous data. My contributions focus on video and 3D-model analysis.
Budget-aware speech recognition with self-supervised representations.
Quality, memory, and latency trade-offs on Pythia-70M.
A local tool for blind and model-assisted structure annotation.
Research software, course experiments, and scientific tools.
Research and internship experience.
Intern · Shanghai Office
Designing enterprise workflow tasks and evaluation datasets, and developing metrics and an evaluation framework for LLMs and agents.
Undergraduate Researcher · Agentic Time Machine
Evaluated agents equipped with Time Machine infrastructure on future-event forecasting tasks.
Undergraduate Researcher · SetupX
Developed an autonomous repository-setup agent with retrieval-augmented experience reuse and refined its agent–environment interaction loop.
Undergraduate Researcher · Data-Analyst Survey
Authored the video and 3D-model analysis sections and curated the accompanying research repository.
Shanghai Jiao Tong University · Artificial Intelligence
I am an undergraduate at the School of Artificial Intelligence, Shanghai Jiao Tong University, and an intern at Microsoft Research Asia.
I work on self-evolving LLM agents, tool learning, and agent evaluation. My research includes automated repository setup, future-event forecasting, and multimodal data analysis.
National Scholarship recipient · GPA 4.04 / 4.3