BadReward: Clean-Label Poisoning of Reward Models in Text-to-Image RLHF

Documenti analoghi

• Poison Once, Control Anywhere: Clean-Text Visual Backdoors in VLM-based Mobile Agents
  di: Wang, Xuan, et al.
  Pubblicazione: (2025)
• BadSampler: Harnessing the Power of Catastrophic Forgetting to Poison Byzantine-robust Federated Learning
  di: Liu, Yi, et al.
  Pubblicazione: (2024)
• BadSkill: Backdoor Attacks on Agent Skills via Model-in-Skill Poisoning
  di: Tie, Guiyao, et al.
  Pubblicazione: (2026)
• Eguard: Defending LLM Embeddings Against Inversion Attacks via Text Mutual Information Optimization
  di: Liu, Tiantian, et al.
  Pubblicazione: (2024)
• Nightshade: Prompt-Specific Poisoning Attacks on Text-to-Image Generative Models
  di: Shan, Shawn, et al.
  Pubblicazione: (2023)