How OpenAI's GPT-4o mini model uses a safety technique called "instruction hierarchy" to prevent misuse and stop "ignore previous instructions" types of attacks (Kylie Robison/The Verge)

Kylie Robison / The Verge:
How OpenAI's GPT-4o mini model uses a safety technique called “instruction hierarchy” to prevent misuse and stop “ignore previous instructions” types of attacks  —  Have you seen the memes online where someone tells a bot to “ignore all previous instructions” …

from Techmeme https://ift.tt/cioOjDn