发表机构
Apple; Caltech（苹果公司; 加州理工学院）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
本文提出探针引导方法，利用扩散模型冻结内部状态构建引导信号，无需额外前向传播，在连续扩散语言模型上实现无条件生成最先进性能，并揭示自动引导机制需低熵弱模型。
AI 中文摘要
我们提出了一种引导流匹配模型的新方法。我们的方法，称为探针引导，利用现有扩散模型的冻结内部状态来构建引导信号。其工作原理与自动引导类似，但在推理时无需额外的前向传播，并提供了一条可靠的路径，确保弱模型和强模型具有相似的动态。我们将此方法应用于连续扩散语言模型并进行基准测试，在无条件生成方面，探针引导取得了新的最先进性能。当应用于1.7B参数的扩散语言模型时，探针引导在多项选择问答基准上持续改进。利用我们的探针，我们研究了传统的自动引导设置，其中强模型是一个弱检查点，并发现弱模型必须来自训练的低熵区域。这些发现既提供了一种改进扩散语言模型的实用方法，也揭示了目前尚不为人所理解的自动引导背后的实际机制。
英文摘要
We introduce a new method to guide flow matching models. Our approach, which we call probe guidance, uses the frozen internal states of an existing diffusion model to construct a guidance signal. This works using a similar principle as autoguidance, but eliminates the need for an additional forward pass at inference time and provides a reliable path to ensure that the weak and strong model share similar dynamics. We apply and benchmark this method on continuous diffusion language models, where probe guidance sets a new state-of-the-art performance on unconditional generation. When applied to a 1.7B diffusion language model, probe guidance consistently improves on multiple choice question answering benchmarks. Using our probes, we study the traditional autoguidance setting where the strong model is a weak checkpoint, and find that the weak model must come from a low-entropy region of training. These findings both provide a practical way to improve diffusion language models and shed light on the actual mechanism behind autoguidance, which is currently poorly understood.
