Mingjian Jiang

About me

I am a second year M.S. student in Computer Science at Stanford University, advised by Tatsunori Hashimoto. Previously, I received my B.S. degree in Computer Science and Mathematics from the University of Toronto, where I fortunately worked with Chris J. Maddison, Amir-massoud Farahmand and Jimmy Ba.

Besides, I spent the summer in 2024 at Tesla Autopilot as a ML Scientist Intern, where I trained large-scale end-to-end self-driving model which is deployable on cars.

Email: jiangm [at] stanford.edu

LinkedIn / Google Scholar / Github

Research Interest

Generally, I am interested in machine learning, with an emphasis on developing scalable but reliable ML models (primarily language models at current stage). My current work focuses on reliability, uncertainty estimation and alignment, while I'm actively exploring questions about model scalability, and data-centric approaches.
I'm especially interested in understanding what enables effective scaling, developing algorithms to ensure trustworthiness and alignment with human values, and investigating data strategies to enhance model capabilities across different tasks / domains.

Start a Conversation: I'm always eager to discuss research ideas and potential collaborations on any topics above. If you're interested in these areas or would like to have a conversation, feel free to reach out!

Publications

• (Under Review) f-PO: Generalizing Preference Optimization with f-divergence Minimization
  Jiaqi Han*, Mingjian Jiang*, Yuxuan Song, Jure Leskovec, Stefano Ermon, Minkai Xu*
  [TL;DR] [Paper] [Code]
• (NeurIPS 2024, Spotlight) Graph-based Uncertainty Metrics for Long-form Language Model Outputs
  Mingjian Jiang, Yangjun Ruan, Prasanna Sattigeri, Salim Roukos, Tatsunori Hashimoto.
  [TL;DR] [Paper] [Code] [Poster]
• (ICML 2023 Deployable Generative AI) Calibrating language models via augmented prompt ensembles
  Mingjian Jiang*, Yangjun Ruan*, Sicong Huang, Saifei Liao, Silviu Pitis, Roger Grosse, Jimmy Ba.
  [TL;DR] [Paper]
• (ICLR 2023 Trustworthy ML) Stealing and Defending Transformer-based Encoders
  Adam Dziedzic, Franziska Boenisch, Mingjian Jiang, Haonan Duan, Nicolas Papernot
  [TL;DR] [Paper]

Selected Awards & Honors

Internship Experience

• Machine Learning Scientist Internship at Tesla Autopilot (Palo Alto, CA) Summer 2024
• Research Internship at Vector Institute (Toronto, Canada) Winter 2022
• Research Internship at Vector Institute (Toronto, Canada) Summer 2021