AIBullisharXiv – CS AI · Mar 46/104
🧠
Conditioned Activation Transport for T2I Safety Steering
Researchers introduce Conditioned Activation Transport (CAT), a new framework to prevent text-to-image AI models from generating unsafe content while preserving image quality for legitimate prompts. The method uses a geometry-based conditioning mechanism and nonlinear transport maps, validated on Z-Image and Infinity architectures with significantly reduced attack success rates.