This issue of AI Deep Research is mapped and its interactive training ground is live — the full written pages are still being drawn. Open the interactive lesson to experiment with rmsnorm vs layernorm · pre-norm vs post-norm in a live sandbox right now.