About me
Hello! I am now an Applied Scientist in Amazon Rufus (Palo Alto, US).
I got a Ph.D. from School of Information, University of Michigan, advised by Prof. VG Vinod Vydiswaran. I also worked with Prof. Qiaozhu Mei between 2015-2020, and Prof. Daniel M. Romero between 2018-2020.
Prior to UM, I received my honored bachelor degree in Computer Science from Chu Kochen Honors College, Zhejiang University in 2017. During my undergraduate period, I was supervised by Prof. Fei Wu.
I primarily working on Natural Language Processing and Machine Learning. My recent research focuses on below topics:
LLM Pre-Training.
LLM Post-Training (Instruction Fine-tuning & RLHF).
LLM Reasoning & Thinking.
Self-Supervised Representation Learning (e.g., contrastive learning).
Robustness of ML models (backdoor attack and defense of LLMs).
Diffusion Models for Non-Autoregressive Generation.
News
[May 2025] Our paper UniConv: Unifying Retrieval and Response Generation for Large Language Model in Conversation is accepted by ACL 2025.
[Sep 2024] Our paper Divide-or-Conquer? Which Part Should You Distill Your LLM? is accepted by EMNLP 2024.
[May 2024] Joined the Amazon Rufus team.
[May 2024] Got married and hosted our wedding!
[Apr 2024] Defended my thesis and left Ann Arbor, a city where I lived for over 8 years.
[Feb 2024] Our paper Divide-or-Conquer? Which Part Should You Distill Your LLM? is now available for public access.
[Oct 2023] Our paper HiCL: Hierarchical Contrastive Learning of Unsupervised Sentence Embeddings is accepted by EMNLP 2023.
[Sep 2023] Our paper PLANNER: Generating Diversified Paragraph via Latent Language Diffusion Model is accepted by NeurIPS 2023.
[Aug 2023] Finished my internship at Machine Learning Research at Apple.