这项研究解决了生成模型版权保护的核心难题——水印不可篡改且不影响生成质量，做AI安全或模型部署的团队值得关注，可以直接参考其黑盒检测方案。
研究人员提出了一种新的生成模型水印方法，将水印直接嵌入到Flow Matching模型的连续动态（速度场）中，而非模型权重或输出。该方法通过随机编码在训练时添加密钥相关的扰动，检测时从黑盒查询中恢复信息，且不改变生成分布。在MNIST和CIFAR-10上的实验表明，该方法能可靠恢复水印，保持生成质量，且无密钥时解码准确率仅为随机水平。
Dynamics-Level Watermarking of Flow Matching Models with Random Codes
We introduce a dynamics-level approach to watermarking generative models. Rather than embedding signals into model weights or outputs, we embed the watermark directly into the learned continuous dynamics -- the velocity field of a flow matching model. We formulate this as random coding over a continuous channel: a key-dependent perturbation is added during training, and the message is recovered at detection time from black-box queries. The perturbation is designed to leave the generated distribution unchanged. Experiments on MNIST and CIFAR-10 across different architectures confirm reliable message recovery, preserved generation quality, and chance-level decoding accuracy without the secret key.