A new study from Google AI Research analyzes how diffusion models generate creative outputs. The team identifies specific mechanisms that drive visual novelty during the denoising process. This technical deep-dive provides a framework for controlling artistic style. Developers can now better tune model randomness to balance fidelity and originality in image generation.