Layer-Aware Representation Filtering: Purifying Finetuning Data to Preserve LLM Safety Alignment

Documenti analoghi

• HarmRLVR: Weaponizing Verifiable Rewards for Harmful LLM Alignment
  di: Liu, Yuexiao, et al.
  Pubblicazione: (2025)
• Be Your Own Red Teamer: Safety Alignment via Self-Play and Reflective Experience Replay
  di: Wang, Hao, et al.
  Pubblicazione: (2026)
• Collaborative Shadows: Distributed Backdoor Attacks in LLM-Based Multi-Agent Systems
  di: Zhu, Pengyu, et al.
  Pubblicazione: (2025)
• Adversarial Attack-Defense Co-Evolution for LLM Safety Alignment via Tree-Group Dual-Aware Search and Optimization
  di: Li, Xurui, et al.
  Pubblicazione: (2025)
• Safety Layers in Aligned Large Language Models: The Key to LLM Security
  di: Li, Shen, et al.
  Pubblicazione: (2024)