Alignment whack-a-mole: Finetuning activates recall of copyrighted books in LLMs

[reconnecting]

Article URL: GitHub - cauchy221/Alignment-Whack-a-Mole-Code: The official code repo of Alignment Whack-a-Mole: Finetuning Activates Verbatim Recall of Copyrighted Books in Large Language Models

Comments URL: Alignment whack-a-mole: Finetuning activates recall of copyrighted books in LLMs | Hacker News

Points: 92

# Comments: 53

Continue reading...