Show HN: A tiny C program where an LLM rewires its DAG while running

1 comments

• youknowme123 13 hours ago ago

  This is nice, it reminds me of earlier work where DAG-based optimization was used to discover effective multi-agent architectures under supervised learning.