扩散模型如何做好可控生成？基于奖励引导的控制生成用于扩散模型中的推理时对齐：教程与综述（英文版）

This tutorial provides an in-depth guide on inference-time guidance and alignment methodsfor optimizing downstream reward functions in diffusion models. While diffusion models arerenowned for their generative modeling capabilities, practical applications in fields such asbiology often require sample generation that maximizes specific metrics (e.g., stability, affinityin proteins, closeness to target structures). In these scenarios, diffusion models can be adaptednot only to generate realistic