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

Documenti analoghi

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