Aaron Branson Cigres Li

E-mail: abcli # connect dot ust dot hk

I am currently a Junior Research Assistant at the MoE Key Laboratory of High Confidence Software Technologies, The Chinese University of Hong Kong. My research interest lies in mechanistic interpretability.

I received my BSc in Mathematics (Statistics Track) from the Hong Kong University of Science and Technology. During my undergraduate studies, I was fortunate to receive guidance from Prof. Yangqiu Song (HKUST), Prof. Pasquale Minervini (UoE), and Prof. Can Yang (HKUST).

I aim to understand how LLMs work, explain how their capabilities emerge, identify why they fail, and use these insights to guide LLM model development and monitoring.

I’ll be in Budapest for EMNLP 2026! If you’re attending, I’d love to meet and chat. Feel free to reach out.

News

Publications

  1. EMNLP 2026 poster: Can Retrieval Heads See Images? EMNLP
    Can Retrieval Heads See Images? Multimodal Retrieval Heads in Long-Context Vision-Language Models (MMRetHeads)
    In The 2026 Conference on Empirical Methods in Natural Language Processing

Education

Hong Kong University of Science and Technology

BSc in Mathematics (Statistics Track)

Sep 2022 – Aug 2026

A special thank you

A special thank you to my girlfriend for her love and support. ❤️

@whois.yvonne on Instagram ↗