A global workspace in language models

A sleep-like consolidation mechanism for LLMs

Language Models Trained on State Media Sources Launder Propaganda

LLMorphism: When humans come to see themselves as language models