the.ai

Interpretability / Circuits

verified

Induction Head

A pair of attention heads that together do something simple and powerful: find where the current token appeared before, and predict whatever followed it last time. It is copying with a lookup. It also appears abruptly during training rather than gradually, at the same moment a model's in-context learning ability jumps.

Viz primitive · attention-heatmaptokens = 12 · heads = 1
query ↓key →

12 queries against 12 keys; a brighter cell means more of that query's attention went to that key. Each query sees itself and everything before it, and nothing after — the upper triangle is masked.

One head's attention over a causal window, under a penalty on distance. Drag the penalty up to watch the head collapse onto the token it just passed — an induction head is the opposite of this, ignoring distance to land on a match far behind it.

0.4

Reviewed by opendroid · 2026-08-18