Are modern large language models (LLMs) vulnerable to privacy attacks that can determine if given data was used for training? Models and dataset are quite large, what should we even expect? Our new paper looks into this exact question. 🧵 (1/10)

New work from my team on identifying memorized training samples for free (no extra computational cost 💵)!

Presenting two papers at the MemFM workshop at ICML! Both touch upon how near duplicates (and beyond) in LLM training data contribute to memorization. - arxiv.org/pdf/2405.15523 - arxiv.org/pdf/2506.20481 Igor Shilov Yves-A. de Montjoye