Part of the success of diffusion models stems from their ability to perform iterative refinement, i.e., repeatedly correcting outputs during generation. However, modern masked discrete diffusion lacks this capability: when a token is generated, it cannot be updated again, even when it introduces an error. Here, we address this limitation by introducing the remasking diffusion model (ReMDM) sampler, a method that can be applied to pretrained masked diffusion models in a principled way and that is derived from a discrete diffusion model with a custom remasking backward process. Most interestingly, ReMDM endows discrete diffusion with a form of inference-time compute scaling. By increasing the number of sampling steps, ReMDM generates natural language outputs that approach the quality of autoregressive models, whereas when the computation budget is limited, ReMDM better maintains quality. ReMDM also improves sample quality of masked diffusion models for discretized images, and in scientific domains such as molecule design, ReMDM facilitates diffusion guidance and pushes the Pareto frontier of controllability relative to classical masking and uniform noise diffusion. We provide the code along with a blog post on the project page: https://guanghanwang.com/remdm

翻译：扩散模型的部分成功源于其执行迭代细化的能力，即能够在生成过程中反复修正输出。然而，现代的掩码离散扩散模型缺乏这种能力：当一个词元被生成后，即使它引入了错误，也无法再次更新。在此，我们通过引入重掩码扩散模型（ReMDM）采样器来解决这一局限性。该方法可以以一种原则性的方式应用于预训练的掩码扩散模型，并且源自一个具有自定义重掩码反向过程的离散扩散模型。最有趣的是，ReMDM 赋予了离散扩散模型一种推理时计算缩放的能力。通过增加采样步数，ReMDM 生成的自然语言输出质量接近自回归模型；而当计算预算有限时，ReMDM 能更好地保持质量。ReMDM 也提升了掩码扩散模型在离散化图像上的样本质量，并且在分子设计等科学领域，ReMDM 促进了扩散引导，并相对于经典的掩码和均匀噪声扩散，推进了可控性的帕累托前沿。我们在项目页面提供了代码和博客文章：https://guanghanwang.com/remdm