Chain of Risk: Safety Failures in Large Reasoning Models and Mitigation via Adaptive Multi-Principle Steering

Documenti analoghi

• Muon-OGD: Muon-based Spectral Orthogonal Gradient Projection for LLM Continual Learning
  di: Lu, Binghang, et al.
  Pubblicazione: (2026)
• When Think-with-Image Meets Safety: What Determines Multimodal Jailbreak Robustness?
  di: Tian, Yuan, et al.
  Pubblicazione: (2026)
• MedGUIDE: Benchmarking Clinical Decision-Making in Large Language Models
  di: Li, Xiaomin, et al.
  Pubblicazione: (2025)
• Dual Identities, Singular Journeys: How Working Tourists' Role Perception Shapes Place Attachment Through Memorable Tourism Experience
  di: Yelin Fang, et al.
  Pubblicazione: (2026)
• Steering Multimodal Large Language Models Decoding for Context-Aware Safety
  di: Liu, Zheyuan, et al.
  Pubblicazione: (2025)