Tianle Gu

Tianle Gu

(M)LLM Safety Evaluation and Alignment, Watermark, Unlearning

About Me

I am currently a master’s student in Big data Engineering at Tsinghua University, advised by Prof. Yujiu Yang, and a research intern at Shanghai Artificial Inetlligence Laboratory. I received my Bachelor’s degree in Computer Science and Techonology from Hunan University.

My research foucses on the safety, alignment, and interpretability of large (multimodal) language models. I have worked on topics such as MLLMs Safety Evaluation, LLM Unlearning, and LLM Watermarking.

My long-term vision is to build LLMs that are more trustworthy, transparent, and secure. I hope to bridge the gap between research and deployment, enabling AI safety techiniques to be widely applied in real-world systems.

.

🎓 Educations

  • 2019.09 - 2023.06, Computer Science and Technology, Hunan University, Changsha China.
  • 2023.09 - 2026.06, Big Data Engineering, Tsinghua University, Shenzhen China.

📝 Publications

Neurips 2024
sym
  • *Tianle Gu, Zeyang Zhou, Kexin Huang, Dandan Liang, Yixu Wang, Haiquan Zhao, Yuanqi Yao, Xingge Qiao, Keqing Wang, Yujiu Yang, Yan Teng, Yu Qiao, Yingchun Wang. MLLMGUARD: A Multi-dimensional Safety Evaluation Suite for Multimodal Large Language Models
    [Paper] [Code]
ACL 2025 Findings
sym
  • *Tianle Gu, *Kexin Huang, Ruilin Luo, Yuanqi Yao, Xiuying Chen, Yujiu Yang, Yan Teng, Yingchun Wang. From Evasion to Concealment: Stealthy Knowledge Unlearning for LLMs
    [Paper] [Code]
Under Review
sym
  • *Tianle Gu, *Zongqi Wang, Kexin Huang, Yuanqi Yao, Xiangliang Zhang, Yujiu Yang, Xiuying Chen. Invisible Entropy: Towards Safe and Efficient Low-Entropy LLM Watermarking [Paper] [Code]
ACL 2025 Main
sym
  • *Zongqi Wang, *Tianle Gu, Baoyuan Wu, Yujiu Yang. MorphMark: Flexible Adaptive Watermarking for Large Language Models [Paper] [Code]
arXiv, Citation 33, Dream begins
sym
  • *Ruilin Luo , *Tianle Gu, *Haoling Li, Junzhe Li, Zicheng Lin, Jiayi Li, Yujiu Yang†. Chain of History: Learning and Forecasting with LLMs for Temporal Knowledge Graph Completion [Paper]
AAAI 2025
sym
  • *Yixu Wang, Tianle Gu, Yan Teng, Yingchun Wang, Xingjun Ma. HoneypotNet: Blackdoor Attacks Against Model Extraction [Paper]

* denotes equal contribution.

Projects

Stars 415
sym
  • Kai Lv, Shuo Zhang, Tianle Gu, Shuhao Xing, Jiawei Hong, Keyu Chen, Xiaoran Liu, Yuqing Yang, Honglin Guo, Tengxiao Liu, Yu Sun, Qipeng Guo, Hang Yan, Xipeng Qiu, CoLLiE: Collaborative Training of Large Language Models in an Efficient Way [Docs] [Github] [Paper]
Stars 86
sym
  • ChatZoo, a lightweight evaluation tool. [Github]

🏅 Honors and Awards

    • 2021*, Hunan Province IoT Competition, 1st Prize
    • 2021 - 2022*, National Scholarship
    • 2019 - 2020*, National Scholarship

💬 Conferences

  • ACL 2025, Austria

🐱 Pets

Egg Dumplings

Dan Jiao

Contact me via Wechat

🏭 Internships

  • 2025.03 - Present, Trustworthy AI Intern, Shanghai Artificial Intelligence Laboratory, Shanghai China
  • 2024.01 - 2024.10, MLLM Evaluation Intern, Shanghai Artificial Intelligence Laboratory, Shanghai China
  • 2023.05 - 2023.08, LLM Distributed Training Engineer (Intern), Shanghai Artificial Intelligence Laboratory, Shanghai China