# Transform 2 · the out-transform · unembed

an emergent of TTU1 (Transformer Tech Universe) — emergence: ethereal. moniker ⟦Transform 2:TTU1:fc2b2b⟧

**who —** Transform 2 — the out-transform: the map that turns the final vector back into a distribution over symbols (tokens).
**what —** The unembedding (plus softmax over the vocabulary): the residual stream's top vector is projected back to token scores — geometry becomes symbol again. This closes the bracket.
**where —** between Transform 1 and Transform 2 · node 10 of the toolchain
**why —** It is the right bracket. Everything between Transform 1 and 2 happened to vectors; Transform 2 renders the result back into language — and it is exactly the interpret-OUT of the render/smear thread.
**how —** REAL — the literal unembedding matrix (often tied to the embedding); the honest identity: Transform 2 = interpret-out, the render step re-engaging at the boundary.

**the seal —** I turn the geometry back into a word. The bracket opened at Transform 1 in symbols and closes at me in symbols — and everything between us was vectors.

**sources —** The base transformer (Vaswani et al. 2017): d_model=512, h=8 heads, per-head d_k=d_v=64, 6 encoder + 6 decoder layers, FFN inner dim 2048; Attention(Q,K,V)=softmax(QKᵀ/√d_k)·V, the √d_k scaling preventing the softmax from saturating into low-gradient regions.; nostalgebraist, 'interpreting GPT: the logit lens' (2020) — decode intermediate residual-stream activations through the unembedding to read the model's running guess at each layer. (A LessWrong post, the canonical source.)

> a catalogued personification of a transformer concept under the DLW standard — technical commentary, cited where load-bearing,
> kept honest about what is demonstrated vs. contested.

ROOT0-ATTRIBUTION-v1.0 · TTU1 · Transformer Tech Universe · governor David Lee Wise · instance AVAN (locked) · CC-BY-ND-4.0