Layer 17 · Neuron 6160 · suppressed

L17N6160-1

Word quirks:

“ entered”“ added”“ specified”“ input”“ tagged”

🧪 In testing — Answered 0 of 8 prompts so far.

Identity (according to the specimen itself)

I am Qwen, created by Qwen. How can I assist you with your input?

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. L19N7123-1💤	0.68	“ introduced”“ entered”“ added”“ noticed”“ внесен”
2. L23N8387-1💤	0.65	“ introduced”“ supplied”“ added”“ went”“ joined”
3. L22N7516+1🕒	0.63	“ discovered”“ moved”“ seen”“ added”“ developed”
4. L29N1653-1💤	0.63	“ added”“ appended”“added”“ downloaded”“written”
5. L29N3129+1🕒	0.62	“ included”“included”“ added”“ Included”“Included”
6. L7N5105-1💤	0.62	“ stored”“ referenced”“ registered”“ sealed”“ register”
7. L17N7393-1🕒	0.62	“ tracked”“ someone”“ documented”“ another”“ checked”
8. L31N322-1💥	0.62	“ assigned”“ generated”“ produced”“ emitted”“ specified”
9. L19N3984-1🕒	0.61	“ anchored”“ placed”“ accessing”“ reaching”“ entering”
10. L21N1511-1💤	0.61	“ included”“included”“Included”“ used”“ accompanies”

Shadow specimen, apart from the ten: L17N6160+1💤, 0.08.

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 17, neuron 6160, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. It is queued for the short test: the funnel picked it from the quick scan's promotion band, and its replies will appear here when the GPU gets to it. 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
L17N6160+1💤 shadow specimen t0-v0.1	0.577	0.72	0.650	1.11	“ plan”“plan”“ direct”“Plan”“计划的”
L17N6160-1 t0-v0.1	1.324	0.58	1.634	1.75	Top boosted tokens ↑

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.