HN Companion◀︎ back | HN Companion home | new | best | ask | show | jobs
Show HN: A tiny C program where an LLM rewires its DAG while running (github.com/kouhxp)
11 points by mrkn1 11 hours ago | 1 comment


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