Lignende værker

• Efficient Adversarial Training in LLMs with Continuous Attacks
  af: Xhonneux, Sophie, et al.
  Udgivet: (2024)
• Adversarial Alignment for LLMs Requires Simpler, Reproducible, and More Measurable Objectives
  af: Schwinn, Leo, et al.
  Udgivet: (2025)
• Model Collapse Is Not a Bug but a Feature in Machine Unlearning for LLMs
  af: Scholten, Yan, et al.
  Udgivet: (2025)
• A Generative Approach to LLM Harmfulness Mitigation with Red Flag Tokens
  af: Dobre, David, et al.
  Udgivet: (2025)
• LLM-Safety Evaluations Lack Robustness
  af: Beyer, Tim, et al.
  Udgivet: (2025)