ConceptSystems & CS
Chain of thought (COT) monitoring
Monitoring an LLM's chain of thought to detect exploits and misbehavior before actions are executed.
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 models1Prompting1
Lecturers
1Connections
2The lectures that define it or assume it, and the concepts next to it.