SafetyPage

Evaluators and safety organisations · Article · Apollo Research

A scalable monitoring research agenda

Read at apolloresearch.ai ↗Read at sourceNo licence statement found

Outline · 38 paragraphs

  1. A scalable monitoring research agenda¶2
  2. Data¶8
  3. Techniques¶15
  4. Basic techniques¶17
  5. Red-teaming¶26
  6. Adversarial games¶29
  7. Automated research¶32