-
Most breakthroughs in deep learning — from simple neural networks to large language models —…
12 min read -
Neural Networks Are Blurry, Symbolic Systems Are Fragmented. Sparse Autoencoders Help Us Combine Them.
Deep LearningNeural and symbolic models compress the world in fundamentally different ways, and Sparse Autoencoders (SAEs)…
17 min read
