CV
I am interested in generative recommendation, long-context memory for LLM agents, and efficient reasoning and retrieval mechanisms. Across these topics, I usually work in a finding-driven way: starting from concrete behavioral observations, then designing mechanisms that are concise, empirically grounded, and easy to reason about.
Education
- B.S. in Artificial Intelligence, University of Science and Technology of China (USTC), School of the Gifted Young Expected 2028
- GPA: 3.87/4.3 Rank: 14/103
- Selected Core Coursework: Discrete Mathematics (100), Probability and Mathematical Statistics (92), Linear Algebra B(1) (90), Data Structures A (90)
- Selected Self-Directed Coursework: Stanford CS229, CS230, CS224n, MIT 6.S184, Berkeley CS61B
Research Experience
Research Intern, Alpha-Lab, USTC
Nov 2025 -- Present
Advisor: Prof. An Zhang
Adaptive-Thinking for Generative Recommendation
Studied a consistent behavioral gap between "think" and "not-think" inference in generative recommendation, especially their differences in predictive entropy, popularity bias, and downstream recommendation quality. Based on these findings, developed an adaptive-thinking framework that selectively invokes reasoning only when uncertainty is high, aiming to balance effectiveness and inference cost. Work in preparation for submission to NeurIPS 2026.
Research Intern, USTC
Mar 2026 -- Present
Advisor: Prof. Xiang Wang
QQMem: Hierarchical Query-to-Query Retrieval for Long-Context Agent Memory
Developing a memory retrieval framework for LLM agents motivated by the observation that direct episode retrieval is often semantically brittle in long-context settings. QQMem replaces episode-level matching with query-space alignment, using structured intermediate queries as semantic anchors to support more stable retrieval and grounded generation. Work in preparation for submission to NeurIPS 2026.
Honors and Awards
- Undergraduate Research Opportunities Program (UROP), Research on Generative Recommendation Systems based on Large Language Models, advised by Prof. An Zhang Dec 2025 – Present
- First Prize (Provincial Level), The 17th Chinese Mathematics Competitions (Non-Math Major, top 20) Oct 2025
- Silver Prize, Outstanding Undergraduates Scholarship Sept 2025
- Bronze Prize, Outstanding Student Scholarship Dec 2024
Skills
- Research: Literature review, experimental design, empirical analysis, and end-to-end implementation for machine learning research.
- Programming: Python, C, Java, Shell/Bash.
- ML Frameworks: PyTorch, Hugging Face, vLLM, verl.
- Model Training: Supervised fine-tuning (SFT), preference optimization (RLHF, DPO, RLvR), parameter-efficient tuning (LoRA), inference-time prompting and reasoning.
- Tools: Linux, Git, Conda, tmux, nvitop, Weights & Biases.
- Engineering: Rapid prototyping with LLM-assisted coding workflows and modern web tooling.