@misc{indiciaeafee46fadadc, title = {Language Models May Verbatim Complete Text They Were Not Explicitly Trained On}, author = {Ken Ziyu Liu and Christopher A. Choquette-Choo and Matthew Jagielski and Peter Kairouz and Sanmi Koyejo and Percy Liang and Nicolas Papernot}, year = {2025}, url = {https://arxiv.org/abs/2503.17514}, note = {Source identifier: 2503.17514} }