Mufei Li (李牧非)'s Homepage
PhD Student @ Georgia Institute of Technology

I’m a second-year PhD student in machine learning at Georgia Institute of Technology, advised by Prof. Pan Li. My current research interests include retrieval-augmented generation (RAG) and graph generative models.
Prior to that, I was a software development engineer (SDE) at Amazon Web Services (AWS) Shanghai AI Lab. I received my bachelor’s degree in Honors Math from NYU Shanghai.
News
Jun 12, 2025 | Check our new paper on better leveraging the LLM KV cache for modeling multi-doc interdependencies, Graph-KV: Breaking Sequence via Injecting Structural Biases into Large Language Models, led by Haoyu! |
---|---|
May 27, 2025 | I’m starting a new position as Research Scientist Intern at Meta Ranking and Foundational AI! I’m based in Sunnyvale, California. |
May 19, 2025 | I’m honored to be a notable reviewer for ICLR 2025! |
May 2, 2025 | Underestimated Privacy Risks for Minority Populations in Large Language Model Unlearning has been accepted by ICML 2025! Congratulations to Rongzhe, Eli, and all the co-authors! |
Apr 3, 2025 | LayerDAG has been selected for spotlight presentation at ICLR 2025 workshop “Will Synthetic Data Finally Solve the Data Access Problem?”! |
Selected Publications
-
ICLRSimple is Effective: The Roles of Graphs and Large Language Models in Knowledge-Graph-Based Retrieval-Augmented GenerationIn International Conference on Learning Representations, 2025Mufei Li and Siqi Miao contributed equally to this work. This work was selected for oral presentation at [ICLR 2025 workshop on Foundation Models in the Wild](https://fm-wild-community.github.io/)
-
ICLR SpotlightLayerDAG: A Layerwise Autoregressive Diffusion Model for Directed Acyclic Graph GenerationIn International Conference on Learning Representations, 2025Selected for spotlight presentation at the ICLR 2025 workshop ["Will Synthetic Data Finally Solve the Data Access Problem?"](https://synthetic-data-iclr.github.io/)