Placeholder: notes on induction heads

This is a filler post so the writing index and home page demonstrate how a list of several posts looks. Delete it once you have real writing.

A real post here might walk through what induction heads are, why the phase change during training matters, and what the pattern $[A][B] \ldots [A] \to [B]$ actually looks like in attention maps — with a figure or an interactive island doing the heavy lifting.