ConceptAI & ML
Jailbreaking LLMs
Prompt engineering techniques used to bypass safety guardrails of large language models by altering the context, such as framing harmful requests as fictional movie scenarios.
Know first
Named as a prerequisite by the lectures that define this. The number is how many say so. A grey one has no page yet.
Large language models1Safety guardrails1
Opens up
Concepts that name this one as a prerequisite.
Lecturers
1Lectures that assume it
1Connections
4The lectures that define it or assume it, and the concepts next to it.