Organization of a Latent Space structure in VAE/GAN trained by navigation data
This research provides insights into how latent spaces in generative models self-organize to reflect data proximity, potentially offering a mechanism for spatial representation in cognition. It is an incremental step in understanding generative models for cognitive mapping.
This paper explores the organization of latent space in a VAE/GAN trained on navigation data, demonstrating that the distance between predicted images is reflected in the distance of their corresponding latent vectors. The system can internally generate temporal sequences, with VAE producing near-accurate replays and GAN introducing instability and novelty.
We present a novel artificial cognitive mapping system using generative deep neural networks, called variational autoencoder/generative adversarial network (VAE/GAN), which can map input images to latent vectors and generate temporal sequences internally. The results show that the distance of the predicted image is reflected in the distance of the corresponding latent vector after training. This indicates that the latent space is self-organized to reflect the proximity structure of the dataset and may provide a mechanism through which many aspects of cognition are spatially represented. The present study allows the network to internally generate temporal sequences that are analogous to the hippocampal replay/pre-play ability, where VAE produces only near-accurate replays of past experiences, but by introducing GANs, the generated sequences are coupled with instability and novelty.