I am a Ph.D. student in Computer Science at UCSD advised by Jingbo Shang. Previously, I was an AI resident at Google Research. I received my B.A. in Computational Linguistics from Peking University.

I pursued research in Natural Language Processing (NLP) because of my multidisciplinary education in both humanities and computer science. My goal is to develop cutting-edge language technologies that are deeply attuned to human behaviors and values, which I refer to as human-centric AI. This involves: (1) language capacity, (2) reasoning ability, and (3) human values. I am honored to be selected as a DeepMind Scholar.

Education

• University of California, San Diego
  Sep. 2021 - Present
  Ph.D. in Computer Science

• The Department of Chinese Language and Literature, Peking University
  Sept. 2015 - Jul. 2019
  B.A. in Computational Linguistics

Experience

• GenAI Team, Meta
  Jun. 2024 - Sep. 2024
  Research Scientist Intern

• Google DeepMind
  Jun. 2022 - Jun. 2024
  Student Researcher

• NLU Team, Google Research
  Oct. 2019 - Sep. 2021
  AI Resident

• Microsoft Research Asia
  Apr. 2019 - Jul. 2019
  Research Intern

Selected Publication

• Retrieval-Augmented Parsing for Complex Graphs by Exploiting Structure and Uncertainty
  Zi Lin, Quan Yuan, Panupong Pasupat, Jeremiah Zhe Liu, Jingbo Shang
  EMNLP findings 2023
  Paper | Code

• ToxicChat: Unveiling Hidden Challenges of Toxicity Detection in Real-World User-AI Conversation
  Zi Lin*, Zihan Wang*, Yongqi Tong, Yangkun Wang, Yuxin Guo, Yujia Wang, Jingbo Shang
  EMNLP findings 2023
  Arxiv | Blog | Data

• Vicuna: An Open-Source Chatbot Impressing GPT-4 with 90% ChatGPT Quality
  Wei-Lin Chiang*, Zhuohan Li*, Zi Lin*, Ying Sheng*, Zhanghao Wu*, Hao Zhang*, Lianmin Zheng*, Siyuan Zhuang*, Yonghao Zhuang*, Joseph E. Gonzalez, Ion Stoica, Eric P. Xing
  Blogpost 2023
  Blog | Demo | Code

• On Compositional Uncertainty Quantification for Seq2seq Graph Parsing
  Zi Lin, Du Phan, Panupong Pasupat, Jeremiah Liu, Jingbo Shang
  ICLR 2023
  Paper | Code

• A Simple Approach to Improve Single-Model Deep Uncertainty via Distance-Awareness
  Jeremiah Liu, Shreyas Padhy, Jie Ren, Zi Lin, Yeming Wen, Ghassen Jerfel, Zachary Nado, Jasper Snoek, Dustin Tran, Balaji Lakshminarayanan
  JMLR 2023
  Paper | Arxiv | Code

• Neural-Symbolic Inference for Robust Autoregressive Graph Parsing via Compositional Uncertainty Quantification
  Zi Lin, Jeremiah Liu, Jingbo Shang
  EMNLP 2022
  Paper | Arxiv | Code

• Towards Collaborative Neural-Symbolic Graph Semantic Parsing via Uncertainty
  Zi Lin, Jeremiah Liu, Jingbo Shang
  ACL Findings 2022
  Paper | Slides

• Measuring and Improving Model-Moderator Collaboration using Uncertainty Estimation
  Ian D. Kivlichan*, Zi Lin*, Jeremiah Liu*, Lucy Vasserman
  ACL 2021 workshop on Online Abuse and Harms (WOAH)
  Paper | Slides | Arxiv | Code

• Simple and Principled Uncertainty Estimation with Deterministic Deep Learning via Distance Awareness
  Jeremiah Liu, Zi Lin, Shreyas Padhy, Dustin Tran, Tania Bedrax-Weiss, Balaji Lakshminarayanan
  Neurips 2020
  Paper | Poster | Arxiv | Code

• Comparing Knowledge-Intensive and Data-Intensive Models for English Resource Semantic Parsing
  Junjie Cao*, Zi Lin*, Weiwei Sun, Xiaojun Wan
  Computational Linguistics 47 (1), 43-68 (also presented in EACL 2021)
  Paper | Slides | Video | Arxiv

• Parsing Meaning Representations: is Easier Always Better?
  Zi Lin, Nianwen Xue
  ACL 2019 Workshop on Designing Meaning Representations (DMR)
  Paper | Slides

• Semantic Role Labeling for Learner Chinese: the Importance of Syntactic Parsing and L2-L1 Parallel Data
  Zi Lin, Yuguang Duan, Yuanyuan Zhao, Weiwei Sun, Xiaojun Wan
  EMNLP 2018
  Paper | Slides | Video | Arxiv | Data

Miscellany

• I learned western painting for nearly ten years and I still enjoy painting as an amateur, including sketching, acrylic painting and digital painting.