Layer 6 · Neuron 4224 · boosted

L6N4224+1

Word quirks:

“ lists”“ list”“ listing”“ List”“list”“ listings”“ Lists”“lists”

🕒 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 for the 2024 list of 100 companies.

Here is a sample of some of the companies:

Company A: 100 companies
Company B: 100 companies
Company C: 100 companies
Company D: 100 companies
Company E: 100 companies
Company F: 100 companies
Company G: 100 companies
Company H: 100

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. L3N3776-1🕒	0.98	“ list”“ lists”“ List”
2. L19N2136+1💤	0.97	“ list”“ listing”“ listed”
3. L2N3151-1💥	0.97	“<\|im_end\|>”“ list”“ lists”
4. L19N4756-1💤	0.97	“ lists”“lists”“ list”
5. L0N6786-1💥	0.97	“ list”“ List”“List”
6. L23N4607-1🕒	0.97	“ lists”“ list”“ Lists”
7. L25N54+1🕒	0.97	“ listing”“Listing”“ list”
8. L19N4605+1🕒	0.96	“ List”“ lists”“ list”
9. L23N7780+1🕒	0.96	“ lists”“ list”“List”
10. L26N836+1🕒	0.96	“ lists”“lists”“ list”

Shadow specimen, apart from the ten: L6N4224-1💤, -0.09.

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 6, neuron 4224, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 2.04, top-1 agreement 0.49) 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
L6N4224+1 t0-v0.1	2.040	0.49	2.367	2.19	Top boosted tokens ↑
L6N4224-1💤 shadow specimen t0-v0.1	0.792	0.69	1.037	1.37	“ '”“ inside”“ your”“ \"”“ sounds”“ basically”“ sound”“ wifi”

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.