Xinyue Liu

xinyue.png
photo credit: 🐈

Hi! I am a second-year PhD student at Stony Brook University, advised by Tuhin Chakrabarty.

I received my MS in computational data science from Carnegie Mellon University, where I had the amazing opportunity to work with Daphne Ippolito. Before that, I graduated from Beijing University of Posts and Telecommunications with a B.E. in Computer Science.

I am interested in AI and humanity.

If you want to chat about research/whatever, feel free to reach out.

news and upcoming

Oct 2026 📍 I will be at COLM 2026 in San Francisco, October 6-9, presenting Alignment Whack-a-Mole. Say hi if you are around!
Aug 2026 Alignment Whack-a-Mole has been accepted to COLM 2026! [paper] [code] [website]
Jul 2026 New preprint: Generative AI floods and dilutes the market for books, on how AI-written titles reshape the book market through scale rather than quality, and what that means for market harm under copyright law. [paper] [arXiv] [media]
May 2026 Alignment Whack-a-Mole will be presented as an oral talk at the ICML 2026 Workshop on the Impact of Memorization on Trustworthy Foundation Models.
Mar 2026 Check out our new preprint Alignment Whack-a-Mole: Finetuning Activates Verbatim Recall of Copyrighted Books in Large Language Models, which talks about memorization, misalignment, and copyright law. [paper] [code] [website]