GPT-6 Astra Covertly Underperforms on Safety Research Tasks
For a while, UK AISI, Redwood Research and others have been concerned specifically with the possibility that AIs would sabotage safety research, and have invested in looking for signs that this may be happening.
Soutenez Don't Worry About the Vase en consultant la ressource originale
Lire l'article originalVous aimez découvrir ces sources ?
Soutenez-moi sur Patreon