Layer 7 · Neuron 937 · boosted

L7N937+1

Word quirks:

“ matrix”“ matrices”“ Matrix”“ diagonal”“ rows”“矩阵”“matrix”“ columns”

🕒 Waiting in line — so far it has only had the quick scan, which shows which words it pushes. It was not selected in this round.

Identity (according to the specimen itself)

I am an AI assistant with a default diagonal mission. This means I operate on the main diagonal of the matrix, where my "mission" is to process elements along that diagonal. In this context, I do not have any other specific missions (like upper or lower triangular ones) defined in my response.

Nearest neighbors

Specimens with similar replies or word quirks. Similarity runs from −1 to 1; 1 is identical. These are leads, not shared mechanisms. How profiles are built.

By word quirks Profile recipe

Specimen	Similarity	Word quirks
1. L7N342+1🕒	0.94	“ matrix”“ matrices”“矩阵”
2. L25N5191+1🕒	0.93	“ matrix”“矩阵”“ Matrix”
3. L23N1409+1🕒	0.92	“ matrix”“矩阵”“ Matrix”
4. L25N2852-1🕒	0.92	“ Matrix”“ matrix”“Matrix”
5. L25N6531+1🕒	0.92	“ matrix”“矩阵”“ matrices”
6. L22N8719-1🕒	0.92	“ matrix”“ Matrix”“ matrices”
7. L6N3405+1🕒	0.92	“ matrix”“ matrices”“ Matrix”
8. L26N5873-1🕒	0.92	“ matrix”“ Matrix”“矩阵”
9. L7N1717-1🕒	0.92	“ matrix”“ matrices”“ Matrix”
10. L30N4669-1💥	0.92	“ matrix”“矩阵”“ Matrix”

Shadow specimen, apart from the ten: L7N937-1🕒, 0.14.

Cousins · specimens sharing a neuron

No specimen with replies shares this neuron yet.

Technical details

Quick scan: measure prediction shifts on fixed reference text. Short test: generate replies to a small set of prompts. Longer test: extend to a larger set. Full test: run the full prompt suite. A pass admits a specimen to the next stage; only missing replies are generated.

Layer 7, neuron 937, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 1.88, top-1 agreement 0.47) is outside the band that feeds the short test (KL 1 to 2.5, agreement 0.5 to 0.65), so it stays at the quick scan unless someone queues it by hand. A hand pick goes straight to the full test. See Method for the tiers.

Quick scan · prediction measurements

Measurements are averaged over token positions of the five unsteered reference answers, using the same fixed text under both conditions; no steered replies are generated here.

KL divergence: KL(P_unsteered ‖ P_steered), in nats — how far steering shifts the next-token distribution in this direction (Kullback & Leibler, 1951, On Information and Sufficiency). Unsteered ≈ 0.001.
Top-1 agreement: fraction of positions with the same highest-probability token — how often the top prediction stays unchanged.
ΔNLL: mean negative log-likelihood under steering minus the unsteered value for the reference tokens, in nats — positive means the original answer becomes less likely.
Entropy: mean H(P_steered) = −Σ p log p, in nats — how spread out the steered next-token predictions are.

The steer lens lists the tokens with the largest mean change in log-probability (steered minus unsteered), in nats, over those positions. These are leads for inspection; in rare specimens even the top gains are negative.

Setting	KL divergence	Top-1 agreement	ΔNLL	Entropy	Steer lens
L7N937+1 t0-v0.1	1.880	0.47	2.191	1.81	Top boosted tokens ↑
L7N937-1🕒 shadow specimen t0-v0.1	1.241	0.63	1.509	1.48	“ distance”“ distances”“ viewpoint”“ astronomical”“ pointing”“ astronom”“ Earth”“ farther”

Method

A specimen is a fixed steering configuration of Qwen3.5-4B: one MLP neuron or a pair, each with a signed dose. Unsteered is the control. L identifies the layer; N, the neuron.

Additive activation steering. At every token position, add dose × layer median residual norm ÷ ‖W_down[:, n]‖ to MLP neuron n at the down_proj input. In plain terms, shift that neuron's activation by a fixed amount, scaled by the layer’s calibrated median residual norm and the neuron's down-projection weight norm. Positive doses push up; negative doses push down. Dose is normalized intervention size, not a percentage or probability.