From pixels to planning: scale-free active inference
This work addresses the problem of scalable generative modeling for dynamic systems, offering a method that integrates active inference with hierarchical structures, though it appears incremental in extending existing frameworks.
The paper introduces a discrete state-space model for generative modeling that generalizes partially observed Markov decision processes to include latent paths, enabling active inference and learning in dynamic settings. It demonstrates the model's application across image classification, movie and music generation, and learning Atari-like games.
This paper describes a discrete state-space model -- and accompanying methods -- for generative modelling. This model generalises partially observed Markov decision processes to include paths as latent variables, rendering it suitable for active inference and learning in a dynamic setting. Specifically, we consider deep or hierarchical forms using the renormalisation group. The ensuing renormalising generative models (RGM) can be regarded as discrete homologues of deep convolutional neural networks or continuous state-space models in generalised coordinates of motion. By construction, these scale-invariant models can be used to learn compositionality over space and time, furnishing models of paths or orbits; i.e., events of increasing temporal depth and itinerancy. This technical note illustrates the automatic discovery, learning and deployment of RGMs using a series of applications. We start with image classification and then consider the compression and generation of movies and music. Finally, we apply the same variational principles to the learning of Atari-like games.