Skip to the content.

Llama-3.2-1B — per-head disassembly

RoPE / GQA / RMSNorm. addr=where-to-read (attn bucket) WRITE=copy/transform (OV diag) bind=top QK token binding idioms=behavioral role QK/OV[…]=SAE-feature opcode (SAE layer only)

Operator roles referenced (hyperlinked inline below): duplicate · induction · prevtok. Full raw listing: llama32_1b_disassembly.txt. See the operator catalog for what each role means.

Discovery pass (causal overlay). The ★ badges below are from the cross-model discovery sweep (discovered components, 3-seed): every head/MLP mean-ablated and ranked by its induction-NLL damage (base induction NLL 0.76). A head is flagged ⚠ UNNAMED-candidate when it is load-bearing but matches no catalogued operator — a lead to dossier. 16 unnamed load-bearing here: 0.31, 1.31, 1.29, 0.29, 0.13, 0.28, 0.14, 0.19, 0.18, 0.20, 1.28, 0.25. Only the sweep’s top-ranked components carry a badge (most heads are not individually load-bearing).

First-order, single-component reads (+ the induction idiom); provisional. Each head line: head · ADDR (where it reads) · WRITE (copy/transform) · top content binding · operator role · ★ discovery-pass causal (when load-bearing). Lines like L.MLP.n#### are MLP neurons (the COMPUTE class — n#### is the neuron’s index in that layer’s gated-MLP intermediate dimension, e.g. Gemma-2-2B has 9216/layer), not attention heads; each lists the top read-tokens → write-tokens (the layer’s most salient few).

Layer 0

Layer 1

Layer 2

Layer 3

Layer 4

Layer 5

Layer 6

Layer 7

Layer 8

Layer 9

Layer 10

Layer 11

Layer 12

Layer 13

Layer 14

Layer 15

Generated from the committed listing + discovery sweep by disassembly_pages.py.