Towards Grounding Conceptual Spaces in Neural Representations
This work addresses the challenge of knowledge representation in AI by attempting to ground conceptual spaces in neural networks, but it appears incremental as it builds on existing frameworks without clear breakthroughs.
The paper tackled the problem of grounding conceptual spaces in neural representations by using an InfoGAN to learn latent spaces from unlabeled data, aiming to bridge symbolic and subsymbolic processing, but no concrete results or numbers were reported.
The highly influential framework of conceptual spaces provides a geometric way of representing knowledge. It aims at bridging the gap between symbolic and subsymbolic processing. Instances are represented by points in a high-dimensional space and concepts are represented by convex regions in this space. In this paper, we present our approach towards grounding the dimensions of a conceptual space in latent spaces learned by an InfoGAN from unlabeled data.