Схожие документы

• LLM-Safety Evaluations Lack Robustness
  по: Beyer, Tim, и др.
  Опубликовано: (2025)
• Efficient Adversarial Training in LLMs with Continuous Attacks
  по: Xhonneux, Sophie, и др.
  Опубликовано: (2024)
• In-Context Learning Can Re-learn Forbidden Tasks
  по: Xhonneux, Sophie, и др.
  Опубликовано: (2024)
• A Coin Flip for Safety: LLM Judges Fail to Reliably Measure Adversarial Robustness
  по: Schwinn, Leo, и др.
  Опубликовано: (2026)
• Soft Prompt Threats: Attacking Safety Alignment and Unlearning in Open-Source LLMs through the Embedding Space
  по: Schwinn, Leo, и др.
  Опубликовано: (2024)