1. Presented three papers at ICML 2025, covering privacy auditing of synthetic text and memorization workshop; 2. Received the best paper award at SaTML 2025; 3. Released ChocoLlama, a family of open-source Dutch LLMs based on Llama-2/3; 4. Presented research on document-level membership inference for LLMs at Usenix Security 2024.
Research Experience
1. Data Scientist Intern at Tesla, working on optimal charge/discharge strategies for residential batteries; 2. Senior Data Scientist at McKinsey and Company, developing NLP solutions; 3. Intern at Microsoft Research, further working on LLM privacy.
Education
1. BSc: KU Leuven, Mechanical Engineering; 2. MEng: UC Berkeley, Energy Technology; 3. MSc: Harvard University, Computer Science; 4. PhD: Imperial College London, supervised by Yves-Alexandre de Montjoye.
Background
Research interests include the impact of frontier technological systems on society, particularly in AI privacy. Currently pursuing a PhD at Imperial College London.
Miscellany
Personal interests include energy, exploring the East Coast of the US. Contact: matthieumeeus@gmail.com. Follow him on Google Scholar, Twitter (X), and LinkedIn for updates.