Decoupled Safety Control: A Safety-Control Algorithm for Training-Free Safety Guidance

Key points

• Training-free safety guidance offers a practical inference-time alternative that intervenes during sampling without modifying model weights.1
• This coupling can work when unsafe factors are explicit, but becomes less reliable once harmful and benign semantics are entangled in the current generation state, often leaving residual unsafe content or unnecessarily degrading benign-generation quality or prompt fidelity.1
• We address this problem with Decoupled Safety Control (DSC), a plug-in safety-control algorithm for training-free safety guidance.1
• Experiments on standard safety benchmarks show that DSC improves representative latent- and text-space safeguards, achieving stronger unsafe-content suppression while preserving benign-generation quality and maintaining efficient inference.1

Sources (1)

• [1]Decoupled Safety Control: A Safety-Control Algorithm for Training-Free Safety Guidance
  arXiv (AI, ML, NLP, CV, robotics, multi-agent) · Oct 15, 09:24 AM

  Diffusion models achieve strong performance in text-to-image generation, but they can still produce unsafe content.

  Training-free safety guidance offers a practical inference-time alternative that intervenes during sampling without modifying model weights.

Extractive summary: sentences quoted from the sources.