2 papers
cs.LG2025
Discovering Chunks in Neural Embeddings for Interpretability
Shuchen Wu, Stephan Alaniz, Eric Schulz +1
Understanding neural networks is challenging due to their high-dimensional, interacting components. Inspired by human cognition, which processes complex sensory data by chunking it…
cs.LG2024
Next state prediction gives rise to entangled, yet compositional representations of objects
Tankred Saanum, Luca M. Schulze Buschoff, Peter Dayan +1
Compositional representations are thought to enable humans to generalize across combinatorially vast state spaces. Models with learnable object slots, which encode information abou…