The Transformer Memory Architecture
Viewing transformers as long-term memory, working memory, and associative retrieval — and how Anthropic’s Workspace and J-Lens results fit into a memory-centric picture of the residual stream.
Transformers
Transformer Blocks as Two Hopfield-Style Memory Systems
How attention mixes tokens via a modern Hopfield retrieval over context, while the MLP mixes channels via a Hopfield-like lookup into learned, persistent memories.