Unified Neural Backdoor Removal with Only Few Clean Samples through Unlearning and Relearning

Documenti analoghi

• Layerwise Convergence Fingerprints for Runtime Misbehavior Detection in Large Language Models
  di: Min, Nay Myat, et al.
  Pubblicazione: (2026)
• CORVUS: Red-Teaming Hallucination Detectors via Internal Signal Camouflage in Large Language Models
  di: Min, Nay Myat, et al.
  Pubblicazione: (2026)
• AutoBackdoor: Automating Backdoor Attacks via LLM Agents
  di: Li, Yige, et al.
  Pubblicazione: (2025)
• Backdoor4Good: Benchmarking Beneficial Uses of Backdoors in LLMs
  di: Li, Yige, et al.
  Pubblicazione: (2026)
• Unlearn to Relearn Backdoors: Deferred Backdoor Functionality Attacks on Deep Learning Models
  di: Shin, Jeongjin, et al.
  Pubblicazione: (2024)