发表机构
ByteDance Seed（字节跳动种子项目）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
该研究提出数据空间迭代框架，无需流离散化即可实现少步生成，在类条件ImageNet 256x256数据集上性能优于离散化基线，为快速生成提供了有效替代方案。
AI 中文摘要
流匹配已成为训练高质量生成模型的可扩展范式，但从学习到的概率流中采样需要多次网络评估。蒸馏可将此成本降低至一次或几次评估；然而，单步生成往往会牺牲质量，因此少步生成是实际应用的操作场景。现有的少步方法沿概率流执行迭代计算，因此需要固定的、手动选择的时间步离散化。这种离散化通常是启发式选择的，调优成本高昂；当不同样本或空间位置的细化难度不同时，它也可能具有局限性。我们引入数据空间迭代，这是一种完全消除流离散化的少步生成框架。从噪声开始，一个共享生成器直接在数据空间中细化其预测，每次迭代都经过训练以生成其容量允许的最佳样本。我们的公式与分布匹配蒸馏（DMD）集成，且改动极小，能够在匹配的训练设置下对不同迭代方法进行受控比较。在类条件ImageNet 256x256数据集上，数据空间迭代的性能优于标准离散化基线，且与通过调度搜索选择的变体相当或更好，无需特定调度的训练。这些结果表明，数据空间迭代为快速生成提供了一种简单有效的替代方案，可替代离散化的流空间迭代。
英文摘要
Flow matching has emerged as a scalable paradigm for training high-quality generative models, but sampling from the learned probability flow requires many network evaluations. Distillation can reduce this cost to one or a few evaluations; however, one-step generation often sacrifices quality, making few-step generation the practical operating regime. Existing few-step methods perform their iterative computation along the probability flow and therefore require a fixed, manually chosen timestep discretization. This discretization is often chosen heuristically and is expensive to tune; it may also be restrictive when refinement difficulty differs across samples or spatial locations. We introduce data-space iteration, a few-step generation framework that removes flow discretization altogether. Starting from noise, a shared generator directly refines its prediction in data space, with every iteration trained to produce the best sample permitted by its capacity. Our formulation integrates with distribution matching distillation (DMD) with minimal changes, enabling a controlled comparison between iteration methods under matched training settings. On class-conditional ImageNet 256x256, data-space iteration outperforms standard discretization baselines and matches or improves upon variants selected through schedule search, without requiring schedule-specific training. These results show that data-space iteration provides a simple and effective alternative to discretized flow-space iteration for fast generation.
