迷失在潜在空间：关于物理模拟的潜在扩散模型实证研究

摘要

扩散模型在推理阶段的高昂计算成本阻碍了其作为快速物理模拟器的应用。在图像和视频生成领域，这一计算瓶颈已通过在自编码器的潜在空间而非像素空间进行生成得到解决。本研究中，我们探讨了类似策略能否有效应用于动态系统的模拟，以及其代价如何。我们发现，潜在空间模拟的准确性对广泛的压缩率（高达1000倍）表现出惊人的鲁棒性。我们还证明，基于扩散的模拟器在准确性上持续优于非生成式模型，并通过预测的多样性更好地补偿了不确定性。最后，我们讨论了从架构到优化器等实际设计选择，这些选择对于训练潜在空间模拟器至关重要。

English

The steep computational cost of diffusion models at inference hinders their use as fast physics emulators. In the context of image and video generation, this computational drawback has been addressed by generating in the latent space of an autoencoder instead of the pixel space. In this work, we investigate whether a similar strategy can be effectively applied to the emulation of dynamical systems and at what cost. We find that the accuracy of latent-space emulation is surprisingly robust to a wide range of compression rates (up to 1000x). We also show that diffusion-based emulators are consistently more accurate than non-generative counterparts and compensate for uncertainty in their predictions with greater diversity. Finally, we cover practical design choices, spanning from architectures to optimizers, that we found critical to train latent-space emulators.

迷失在潜在空间：关于物理模拟的潜在扩散模型实证研究

Lost in Latent Space: An Empirical Study of Latent Diffusion Models for Physics Emulation

摘要

Support