Diverse and Effective Red Teaming with Auto-generated Rewards and Multi-step Reinforcement Learning

Ähnliche Einträge

• The Instruction Hierarchy: Training LLMs to Prioritize Privileged Instructions
  von: Wallace, Eric, et al.
  Veröffentlicht: (2024)
• Reinforcement Learning with Conditional Expectation Reward
  von: Xiao, Changyi, et al.
  Veröffentlicht: (2026)
• RedTopic: Toward Topic-Diverse Red Teaming of Large Language Models
  von: Ding, Jiale, et al.
  Veröffentlicht: (2025)
• Gradient Regularization Prevents Reward Hacking in Reinforcement Learning from Human Feedback and Verifiable Rewards
  von: Ackermann, Johannes, et al.
  Veröffentlicht: (2026)
• Tree-based Dialogue Reinforced Policy Optimization for Red-Teaming Attacks
  von: Guo, Ruohao, et al.
  Veröffentlicht: (2025)