跟随均值：参考引导的流匹配

摘要

现有可控生成方法通常依赖于微调、辅助网络或测试时搜索。我们证明流匹配提供了一种不同的控制接口：通过示例进行适应。对于确定性插值，速度场完全由条件端点均值控制；偏移该均值即可改变流本身。这引出了一个简单的可控生成原则：通过改变预训练模型所遵循的参考集来引导其生成。我们以两种形式实现这一思想。参考均值引导无需训练：它从参考库中计算封闭形式的端点均值校正，并将其应用于冻结的FLUX.2-klein（4B）模型，在保持提示词、随机种子和权重不变的情况下，实现对颜色、身份、风格和结构的控制。半参数引导通过显式均值锚点和学习的残差优化器将同一思想进行摊销，在AFHQv2上达到无条件DiT-B/4质量，同时允许在推理时切换参考集。这些结果指向一个更广泛的方向：生成模型通过数据而非参数更新来适应。

English

Existing approaches to controllable generation typically rely on fine-tuning, auxiliary networks, or test-time search. We show that flow matching admits a different control interface: adaptation through examples. For deterministic interpolants, the velocity field is solely governed by a conditional endpoint mean; shifting this mean shifts the flow itself. This yields a simple principle for controllable generation: steer a pretrained model by changing the reference set it follows. We instantiate this idea in two forms. Reference-Mean Guidance is training-free: it computes a closed-form endpoint-mean correction from a reference bank and applies it to a frozen FLUX.2-klein (4B) model, enabling control of color, identity, style, and structure while keeping the prompt, seed, and weights fixed. Semi-Parametric Guidance amortizes the same idea through an explicit mean anchor and learned residual refiner, matching unconditional DiT-B/4 quality on AFHQv2 while allowing the reference set to be swapped at inference time. These results point to a broader direction: generative models that adapt through data, not parameter updates.