Toward a Principled Framework for Agent Safety Measurement

Documenti analoghi

• Toward Principled LLM Safety Testing: Solving the Jailbreak Oracle Problem
  di: Lin, Shuyi, et al.
  Pubblicazione: (2025)
• Text-to-Image Models Leave Identifiable Signatures: Implications for Leaderboard Security
  di: Naseh, Ali, et al.
  Pubblicazione: (2025)
• Exploiting Leaderboards for Large-Scale Distribution of Malicious Models
  di: Suri, Anshuman, et al.
  Pubblicazione: (2025)
• SAGA: A Security Architecture for Governing AI Agentic Systems
  di: Syros, Georgios, et al.
  Pubblicazione: (2025)
• DROP: Poison Dilution via Knowledge Distillation for Federated Learning
  di: Syros, Georgios, et al.
  Pubblicazione: (2025)