Beizhe Hu 胡焙哲

I am a 1st-year PhD candidate under the supervision of Prof. Juan Cao at Institute of Computing Technology, Chinese Academy of Sciences (ICT, CAS). My research interests include:

  • Fake News/Misinformation Detection and Fact-Checking
  • Natural Language Understanding and Data Mining
  • AI Safety and AI for Safety

Contact: hubeizhe21s [AT] ict [dot] ac [dot] cn

GitHub    /    Google Scholar    /    DBLP

News
12/2024 Invited to serve as a PC member for WWW 2025.
07/2024 One co-authored papers got accepted by CIKM 2024.
04/2024 One co-authored papers got accepted by IJCAI 2024.
02/2024 Invited to serve as a Reviewer for ACL Rolling Review.
12/2023 One first-authored paper got accepted by AAAI 2024.
05/2023 One first-authored paper got accepted by ACL 2023.
Publications
Preprint
Understanding News Creation Intents: Frame, Dataset, and Method
Zhengjia Wang, Danding Wang, Qiang Sheng, Juan Cao, Silong Su, Yifan Sun, Beizhe Hu, and Siyuan Ma
arXiv:2312.16490
2024
[CIKM 2024]   Let Silence Speak: Enhancing Fake News Detection with Generated Comments from Large Language Models
Qiong Nan, Qiang Sheng, Juan Cao,Beizhe Hu, Danding Wang, and Jintao Li
Proceedings of the 33rd ACM International Conference on Information and Knowledge Management
Preprint / Paper / GitHub Repo / Chinese Blog
TL;DR: We prompt LLMs to role-play social media users to obtain generated comments for enhancing fake news detection.
[IJCAI 2024]  Ten Words Only Still Help: Improving Black-Box AI-Generated Text Detection via Proxy-Guided Efficient Re-Sampling
Yuhui Shi, Qiang Sheng, Juan Cao, Hao Mi, Beizhe Hu, and Danding Wang
Proceedings of the 33rd International Joint Conference on Artificial Intelligence
Preprint / GitHub Repo / Chinese Blog
TL;DR: To detect and attribute text generated by black-box LMs, we estimate their generation probabilities of representative words guided by a white-box proxy LM to obtain a strong feature.
[AAAI 2024]  Bad Actor, Good Advisor: Exploring the Role of Large Language Models in Fake News Detection
Beizhe Hu, Qiang Sheng, Juan Cao, Yuhui Shi, Yang Li, Danding Wang, and Peng Qi
Proceedings of the 38th AAAI Conference on Artificial Intelligence
Preprint / Paper / GitHub Repo / English Video & Slides
TL;DR: Large LMs generally underperform fine-tuned Small LMs for fake news detection, but they can be good advisors by providing rationales.
2023
[ACL 2023]  Learn over Past, Evolve for Future: Forecasting Temporal Trends for Fake News Detection
Beizhe Hu, Qiang Sheng, Juan Cao, Yongchun Zhu, Danding Wang, Zhengjia Wang, and Zhiwei Jin
Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics
Preprint / Paper / Poster / Slides
TL;DR: We propose to address the temporal shift issue in real-world fake news detection systems via forecasting topic-level trends and accordingly adjusting the detector update strategy.
Academic Services
Conf. Reviewer/PC Member TheWebConf (WWW) 2025

ACL Rolling Review (Feb. 2024~)
Education
2024- Ph.D in Computer Science and Technology (Candidate)
University of Chinese Academy of Sciences (Cultivation Unit: Institute of Computing Technology, Chinese Academy of Sciences)
2021-2024 M. E. in Computer Technology
University of Chinese Academy of Sciences (Cultivation Unit: Institute of Computing Technology, Chinese Academy of Sciences)
2017-2021 B. E. in Software Engineering
School of Computer Science and Technology, Xidian University
Honors and Awards
2024 Pacemaker to Merit Student of UCAS, UCAS
2023 National Scholarship, Ministry of Education of China
2023 First-Class Academic Scholarship, University of Chinese Academy of Sciences
2023 Merit Student, University of Chinese Academy of Sciences
2019, 2020 First-Class Academic Scholarship, Xidian University

Design: Jon Barron, Limin Yang, Qiang Sheng and Yongchun Zhu