From Refusal to Recovery: A Control-Theoretic Approach to Generative AI Guardrails

Documenti analoghi

• Human-AI Safety: A Descendant of Generative AI and Control Systems Safety
  di: Bajcsy, Andrea, et al.
  Pubblicazione: (2024)
• Robots that Learn to Safely Influence via Prediction-Informed Reach-Avoid Dynamic Games
  di: Pandya, Ravi, et al.
  Pubblicazione: (2024)
• MAGICS: Adversarial RL with Minimax Actors Guided by Implicit Critic Stackelberg for Convergent Neural Synthesis of Robot Safety
  di: Wang, Justin, et al.
  Pubblicazione: (2024)
• From Refusal Tokens to Refusal Control: Discovering and Steering Category-Specific Refusal Directions
  di: Alagharu, Rishab, et al.
  Pubblicazione: (2026)
• Toward Reusability of AI Models Using Dynamic Updates of AI Documentation
  di: Bajcsy, Peter, et al.
  Pubblicazione: (2026)