Bits Leaked per Query: Information-Theoretic Bounds on Adversarial Attacks against LLMs

Documenti analoghi

• Self-Evaluation as a Defense Against Adversarial Attacks on LLMs
  di: Brown, Hannah, et al.
  Pubblicazione: (2024)
• MPAT: Building Robust Deep Neural Networks against Textual Adversarial Attacks
  di: Zhang, Fangyuan, et al.
  Pubblicazione: (2024)
• Text-CRS: A Generalized Certified Robustness Framework against Textual Adversarial Attacks
  di: Zhang, Xinyu, et al.
  Pubblicazione: (2023)
• TFL: Targeted Bit-Flip Attack on Large Language Model
  di: Guo, Jingkai, et al.
  Pubblicazione: (2026)
• SBFA: Single Sneaky Bit Flip Attack to Break Large Language Models
  di: Guo, Jingkai, et al.
  Pubblicazione: (2025)