#ai-policy

  1. Alignment Whack-a-Mole: Finetuning Activates Verbatim Recall of Copyrighted Books in LLMs
  2. The Pro-Human AI Declaration