Holographic generative flows with AdS/CFT
This work introduces a novel, physically interpretable approach to generative modeling, potentially impacting machine learning by leveraging quantum gravity concepts, though it is incremental as it builds on existing flow-matching methods.
The authors tackled generative modeling by integrating the AdS/CFT correspondence from quantum gravity into flow-matching algorithms, resulting in faster and higher-quality convergence on toy datasets and MNIST compared to physics-free models.
We present a framework for generative machine learning that leverages the holographic principle of quantum gravity, or to be more precise its manifestation as the anti-de Sitter/conformal field theory (AdS/CFT) correspondence, with techniques for deep learning and transport theory. Our proposal is to represent the flow of data from a base distribution to some learned distribution using the bulk-to-boundary mapping of scalar fields in AdS. In the language of machine learning, we are representing and augmenting the flow-matching algorithm with AdS physics. Using a checkerboard toy dataset and MNIST, we find that our model achieves faster and higher quality convergence than comparable physics-free flow-matching models. Our method provides a physically interpretable version of flow matching. More broadly, it establishes the utility of AdS physics and geometry in the development of novel paradigms in generative modeling.