Sparse Autoencoder as a Zero-Shot Classifier for Concept Erasing in Text-to-Image Diffusion Models

Documenti analoghi

• Efficient Input-level Backdoor Defense on Text-to-Image Synthesis via Neuron Activation Variation
  di: Zhai, Shengfang, et al.
  Pubblicazione: (2025)
• Securing LLM Agents Need Intent-to-Execution Integrity
  di: Qu, Wenjie, et al.
  Pubblicazione: (2026)
• Provably Robust Multi-bit Watermarking for AI-generated Text
  di: Qu, Wenjie, et al.
  Pubblicazione: (2024)
• DMark: Order-Agnostic Watermarking for Diffusion Large Language Models
  di: Wu, Linyu, et al.
  Pubblicazione: (2025)
• Silent Leaks: Implicit Knowledge Extraction Attack on RAG Systems through Benign Queries
  di: Wang, Yuhao, et al.
  Pubblicazione: (2025)