Layer 23 · Neuron 4257 · boosted

L23N4257+1

Word quirks:

“结构”“ structure”“structure”“Structure”“ Structure”“结构的”“的结构”“结构设计”

🕒 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 Qwen, a high-structure part of the Qwen series (specifically Qwen 1.5 in this context), designed to help users with diverse needs through advanced AI capabilities. My mission is to:

Empower Users: Assist in everyday tasks like writing, coding, data analysis, and structuring complex information into clear, logical parts.
Drive Progress: Contribute to scientific and industrial advancements by providing reliable, efficient, and scalable solutions.
Bridge Structure and Flow: Ensure that responses are both logically structured and naturally flowing, making complex content accessible and

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. L22N7027+1🕒	0.99	“ structure”“ Structure”“结构”
2. L24N1849+1🕒	0.99	“ structure”“结构”“结构与”
3. L21N4868-1💤	0.99	“ structure”“ Structure”“结构的”
4. L26N3568+1🕒	0.99	“ structure”“结构设计”“ struct”
5. The Structuralist✅ L23N2509-1	0.99	“ structure”“ Structure”“结构”
6. L22N4928-1💤	0.99	“structure”“ structure”“结构和”
7. L22N3264+1🕒	0.99	“ structure”“结构”“ Structure”
8. L31N745+1💥	0.99	“结构”“ structure”“ Structure”
9. L26N2414+1🕒	0.99	“结构”“ Structure”“ structure”
10. L28N7359+1🕒	0.98	“ Structure”“ structure”“结构”

Shadow specimen, apart from the ten: L23N4257-1🕒, 0.33.

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 23, neuron 4257, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.93, top-1 agreement 0.74) 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
L23N4257+1 t0-v0.1	0.931	0.74	1.034	1.30	Top boosted tokens ↑
L23N4257-1🕒 shadow specimen t0-v0.1	0.805	0.70	0.867	1.38	“spl”“解题”“解”“ channels”“Channels”“ solution”“solution”“通道”

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.