- A Multi-Agent Collaborative Framework for Database Konb Tuning. (SIGMOD'26)
- Hidden Question Representations Tell Non-Factuality Within and Across Large Language Models. (EMNLP'25)
- A Query Optimization Method Utilizing Large Language Models. (arXiv:2503.06902)
- LLMIdxAdvis: Resource-Efficient Index Advisor Utilizing Large Language Model. (arXiv:2503.07884)
- E2ETune: End-to-End Knob Tuning via Fine-tuned Generative Language Model. (VLDB'26)
- OmniSQL: Synthesizing High-quality Text-to-SQL Data at Scale. (VLDB'25)
- Dynamic Scaling of Unit Tests for Code Reward Modeling. (ACL'25)
- VisualSimpleQA: A Benchmark for Decoupled Evaluation of Large Vision-Language Models in Fact-Seeking Question Answering. (arXiv:2503.06492)
- LoRS: Efficient Low-Rank Adaptation for Sparse Large Language Model. (arXiv:2501.08582)
- SAM Decoding: Speculative Decoding via Suffix Automaton. (ACL'25)
- P^2Law: Scaling Law for Post-Training After Model Pruning. (ACL'25)
Research Experience
Professor at the School of Information, Renmin University of China
Education
Ph.D. in Computer Science and Technology, Tsinghua University, supervised by Professor Jie Tang and Professor Juanzi Li
Background
Research Interests: Data mining and knowledge discovery, with a focus on tailoring large language models (LLMs) for structured data processing. Specific research areas include model alignment methods (such as data synthesis and learning from human-AI feedback), model compression techniques, and efficient inference methods.
Miscellany
Personal Interests: Looking for highly-motivated students to work with