Persona Attack: Incremental Memory Injection Jailbreak Attack against Large Language Models

Documenti analoghi

• Beyond Attack Success Rate: Temporal Logit Observability for LLM Safety Failures
  di: Park, Junyoung, et al.
  Pubblicazione: (2026)
• SoK: Robustness in Large Language Models against Jailbreak Attacks
  di: Xu, Feiyue, et al.
  Pubblicazione: (2026)
• On Jailbreaking Quantized Language Models Through Fault Injection Attacks
  di: Zahran, Noureldin, et al.
  Pubblicazione: (2025)
• AISA: Awakening Intrinsic Safety Awareness in Large Language Models against Jailbreak Attacks
  di: Song, Weiming, et al.
  Pubblicazione: (2026)
• Align is not Enough: Multimodal Universal Jailbreak Attack against Multimodal Large Language Models
  di: Wang, Youze, et al.
  Pubblicazione: (2025)