Sparse Autoencoders as Plug-and-Play Firewalls for Adversarial Attack Detection in VLMs

Documenti analoghi

• Very Efficient Listwise Multimodal Reranking for Long Documents
  di: Sun, Yiqun, et al.
  Pubblicazione: (2026)
• Traveling Across Languages: Benchmarking Cross-Lingual Consistency in Multimodal LLMs
  di: Wang, Hao, et al.
  Pubblicazione: (2025)
• Return of Frustratingly Easy Unsupervised Video Domain Adaptation
  di: Wei, Pengfei, et al.
  Pubblicazione: (2026)
• Evaluating Multimodal Large Language Models on Vertically Written Japanese Text
  di: Sasagawa, Keito, et al.
  Pubblicazione: (2025)
• An Image is Worth 1/2 Tokens After Layer 2: Plug-and-Play Inference Acceleration for Large Vision-Language Models
  di: Chen, Liang, et al.
  Pubblicazione: (2024)