I am a PhD student at Northeastern University, advised by Prof. Weiyan Shi. I also work closely with Prof. David Bau. I was a scholar in the MATS program, working on reasoning at Prof. Dawn Song’s Lab. I obtained my Master’s degree in Computer Science at UMass Amherst where I was lucky to be advised by Prof. Andrew McCallum and Prof. Hong Yu. I finished my undergraduate study in Computer Science at HKUST.

I am generally interested in understanding the mechanisms of AI models to improve and control them. I am currently working on post-training, especially on understanding and mitigating emergent behaviors. I am broadly interested in continual learning, reasoning and interp. Feel free to email me if you would like to collaborate.

News

• Sep 2026Gave a talk at Meta on how chat-template tokens shape AI safety, covering The Piggyback Hypothesis and LLMs Encode Harmfulness and Refusal Separately.
• Sep 2026The Piggyback Hypothesis of Generalization was accepted to NeurIPS 2026.
• Sep 2026Our True Thinking Score was added to the Tinker cookbook by Thinking Machines.
• Aug 2026Can Aha Moments Be Fake? was accepted to EMNLP 2026 Findings.
• Jul 2026Quanta Magazine covered our work on true vs. decorative thinking steps in chain-of-thought.
• Jun 2026New preprint: The Piggyback Hypothesis, explaining and mitigating emergent misalignment.

Publications

* Equal contribution

Academic service

Reviewer for ICML, NeurIPS, CoNLL, AAAI, ICLR