Researchers propose focusing on identification of certain cognitive elements in LLMs that indicate when AI systems may take an unwanted action.
Stronger AI Safety Requires Peeking Inside the 'Black Box'
Résumé + lien vers la source. Sans paywall, sans tracking.