LUMIA: Linear probing for Unimodal and MultiModal Membership Inference Attacks leveraging internal LLM states

Ähnliche Einträge

• LPASS: Linear Probes as Stepping Stones for vulnerability detection using compressed LLMs
  von: Ibanez-Lissen, Luis, et al.
  Veröffentlicht: (2025)
• Nob-MIAs: Non-biased Membership Inference Attacks Assessment on Large Language Models with Ex-Post Dataset Construction
  von: Eichler, Cédric, et al.
  Veröffentlicht: (2024)
• MMA-Diffusion: MultiModal Attack on Diffusion Models
  von: Yang, Yijun, et al.
  Veröffentlicht: (2023)
• On the Feasibility of Using MultiModal LLMs to Execute AR Social Engineering Attacks
  von: Bi, Ting, et al.
  Veröffentlicht: (2025)
• Data Provenance Auditing of Fine-Tuned Large Language Models with a Text-Preserving Technique
  von: Li, Yanming, et al.
  Veröffentlicht: (2025)