Posts

Auditing language models for hidden objectives 2025-03-13T19:18:32.638Z
Image Hijacks: Adversarial Images can Control Generative Models at Runtime 2023-09-20T15:23:48.898Z

Comments