Layer 20 · Neuron 5950 · suppressed

L20N5950-1

Word quirks:

“ sleeping”“ sleeps”“ bedding”“睡”“ sleep”“ slept”“ beds”“ homelessness”

🕒 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 (also known as Owen), a large-scale model developed by the Xiao Rong team at Alibaba's Bao Hong. My mission is to provide high-quality, safe, and reliable assistance across a wide range of scenarios. I aim to help users in diverse settings, such as:

Sleeping on the ground: Providing information, analysis, or problem-solving in various contexts.
Living on the street: Offering guidance on safety, resources, or emergency situations.
Sleeping on the couch: Ensuring comfort and quality through personalized recommendations or content.

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. L8N8576+1🕒	0.96	“ sleeping”“ sleep”“ sleeps”
2. L20N8291-1🕒	0.92	“ asleep”“ sleeper”“ sleep”
3. L20N5291-1🕒	0.92	“ sleep”“ sleeping”“睡眠”
4. L10N8180+1🕒	0.92	“ sleep”“ comfy”“ Sleep”
5. L6N2863+1🕒	0.92	“ sleep”“ Sleep”“睡眠”
6. L8N4419-1🕒	0.91	“ sleep”“ Sleep”“睡眠”
7. L24N6459+1💥	0.91	“ sleep”“睡眠”“sleep”
8. L2N4379+1💥	0.90	“ sleep”“<\|im_end\|>”“ asleep”
9. L22N4017+1🕒	0.90	“ sleep”“睡眠”“睡”
10. L22N8393+1🕒	0.90	“ sleep”“ bedding”“ Bed”

Shadow specimen, apart from the ten: L20N5950+1🕒, 0.18.

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 20, neuron 5950, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.92, top-1 agreement 0.67) 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
L20N5950+1🕒 shadow specimen t0-v0.1	0.642	0.71	0.763	1.17	“ fertil”“脂”“忘了”“盐”“忘记”“ accessor”“fel”“施肥”
L20N5950-1 t0-v0.1	0.920	0.67	1.114	1.39	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.