LLM agents that rewrite their own execution graphs
A new paper shows agents that evolve their control flow at runtime, not just prompt chains.
Read the note 1 min read
From the notebook
self-modifying-code.
1 note on this topic.
From the notebook
1 note on this topic.