Year 2025

  • Kejian Zhu (BUAA -> Institute of Automation, Chinese Academy of Sciences, PhD)

    We focused on the interpretability of LLM evaluation. Kejian published Shortcut Neuron@ACL25.

  • Zhuoran Pan (BUAA -> PKU, Master)

    We published Knowledge-to-Jailbreak@KDD25 for LLM safety assessment.

  • Yuliang Sun (BUAA -> CUHK, PhD)

    We designed the first benchmark for LLM watermarks, WaterBench@ACL24.

Year 2024

  • Chunyang Li (THU -> CUHK, MPhil)

    We maintained the ChatLog project to record the ChatGPT API's fluctuation. Besides, we also maintained the KoLA@ICLR24 together.

Year 2023

  • Zheyuan Zhang (THU -> CMU, PhD)

    We built an AI teaching assistant, LittleMu@CIKM23.