Mingjie Li

I am currently a 3rd-year master student (expected graduation time: March 2024) in Computer Science Department at Shanghai Jiao Tong University (SJTU), advised by Prof. Quanshi Zhang. Prior to that, I receieved the Bachelor degree in CS (IEEE honor class) from SJTU.

My current research interest lies in Explainable AI. This topic mainly focuses on designing tools to open the black-box of deep neural networks. I believe XAI could boost AI research. I am also interested in various exciting AI applications.

🔥 I am actively seeking for full-time machine learning engineer opportunities starting in 2024. Please feel free to contact me if interested.

Email  |  Google Scholar  |  Twitter  |  Github

profile photo
News
😶 [2023-06] Refactor this page.
😇 [2021-09] I started my master.
😃 [2021-06] I graduated.
Industrial Experiences
Still under construction, stay tuned ...
Icon Machine learning engineer for recommender system (intern), Bytedance (Douyin), Shanghai, China    2023.3 - 2023.7
Publications
(* means equal contribution)
Does a Neural Network Really Encode Symbolic Concepts?   [ paper  |  知乎  |  机器之心  |  code ]
Mingjie Li, Quanshi Zhang
ICML, 2023
Defining and Quantifying the Emergence of Sparse Concepts in DNNs   [ paper  |  知乎  |  code ]
Jie Ren*, Mingjie Li*, Qirui Chen, Huiqi Deng, Quanshi Zhang
CVPR, 2023
Towards Theoretical Analysis of Transformation Complexity of ReLU DNNs   [ paper  |  code ]
Jie Ren*, Mingjie Li*, Meng Zhou, Shih-Han Chan, Quanshi Zhang
ICML, 2022
Visualizing the Emergence of Intermediate Visual Patterns in DNNs   [ paper  |  知乎  |  video  |  code ]
Mingjie Li, Shaobo Wang, Quanshi Zhang
NeurIPS, 2021
Interpreting and Disentangling Feature Components of Various Complexity from DNNs   [ paper  |  知乎  |  bilibili  |  code ]
Jie Ren*, Mingjie Li*, Zexu Liu, Quanshi Zhang
ICML, 2021
Interpreting and Boosting Dropout from a Game-Theoretic View   [ paper  |  知乎 ]
Hao Zhang, Sen Li, Yinchao Ma, Mingjie Li, Yichen Xie, Quanshi Zhang
ICLR, 2021
Can the Inference Logic of Large Language Models be Disentangled into Symbolic Concepts?   [ arxiv ]
Wen Shen, Lei Cheng, Yuxiao Yang, Mingjie Li, Quanshi Zhang
Preprint, 2023
Technical Note: Defining and Quantifying AND-OR Interactions for Faithful and Concise Explanation of DNNs   [ arxiv ]
Mingjie Li, Quanshi Zhang
Technical Note
Selected Awards
◾ National Scholarship for graduate students, Ministry of Education of China, 2022, 2023
◾ Shanghai outstanding graduates, Shanghai Municipal Education Commission, 2021
◾ National Scholarship for undergraduate students, Ministry of Education of China, 2019, 2020
◾ The Tang Lixin Scholarship, 2018
Academic Services
◾ Conference Reviewer: ICML 2022, CVPR 2023, ICCV 2023, NeurIPS 2023, ICLR 2024
◾ Journal Reviewer: IJCV

Last updated: May 2023.
Source code from Jon Barron's website.