@levanderhart
https://t.co/0UJ18H8Juj
calendar_today18-02-2010 22:07:11
3,9K Tweets
2,2K Followers
518 Following
1 year ago
Based on your current understanding of LLMs, do you believe they are capable of considering saying one thing in response to a prompt, but then deciding to say something else instead? (The model itself, not a separate censor layer)