Tag
A study in Nature Machine Intelligence presents GCML, a neuromorphic model using neural sampling from cognitive maps for goal-directed imagination and planning, enabling energy-efficient, adaptive problem-solving. An educational tool is available for hands-on experimentation.
This paper presents gradCSCG, a differentiable reformulation of the Clone-Structured Causal Graph algorithm for end-to-end learning of cognitive maps directly from image sequences, combining it with a vector-quantized variational autoencoder.