What prompts are used in RLAIF?
#13
by
chujiezheng
- opened
Thanks for your great work. Could you disclose what prompts are used in the policy sampling steps in RLAIF?
Thank you for your question! It's using the same prompt from the Nectar dataset here: https://huggingface.co/datasets/berkeley-nest/Nectar
banghua
changed discussion status to
closed