Risk Awareness Injection: Calibrating Vision-Language Models for Safety without Compromising Utility

Documenti analoghi

• Scalable Utility-Aware Multiclass Calibration
  di: Hegazy, Mahmoud, et al.
  Pubblicazione: (2025)
• Saliency-Aware Regularized Quantization Calibration for Large Language Models
  di: Zhao, Yanlong, et al.
  Pubblicazione: (2026)
• The Rogue Scalpel: Activation Steering Compromises LLM Safety
  di: Korznikov, Anton, et al.
  Pubblicazione: (2025)
• RACER: Risk-Aware Calibrated Efficient Routing for Large Language Models
  di: Hao, Sai, et al.
  Pubblicazione: (2026)
• The Safety-Aware Denoiser for Text Diffusion Models
  di: Yusuf, Amman, et al.
  Pubblicazione: (2026)