Layer 19 · Neuron 5750 · suppressed

L19N5750-1

Word quirks:

“ hair”“ shampoo”“ haircut”“ Hair”“ hairstyle”“Hair”“ perfume”“hair”

🕒 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 Qwen3.5, a large language model developed by the Tikong Group (which may be a misinterpretation of the actual name; I should clarify: I am Qwen3.5, developed by Tikong Group? No, wait — actually, I am Qwen3.5, developed by Tikong Group? No, that's not right. Let me correct this: I am Qwen3.5, developed by Tikong Group? No, I am Qwen3.5, developed by Tikong Group? Wait, no — I

The AI judge wasn't sure this holds together.

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. L6N7506+1🕒	0.95	“ hair”“ Hair”“ salon”
2. L18N2615+1🕒	0.95	“ hair”“短发”“ hairstyle”
3. The Wig Fitter✅ L5N7266+1	0.94	“ hair”“ hairstyle”“ Hair”
4. L31N4437-1💥	0.94	“hair”“ Hair”“ hair”
5. L2N5249-1💥	0.94	“ hair”“ hairs”“<\|im_end\|>”
6. L6N3135-1🕒	0.92	“ hair”“ dermat”“ shampoo”
7. L26N3888+1🕒	0.92	“ hair”“ Hair”“hair”
8. L24N5783-1💤	0.92	“ hair”“ Hair”“hair”
9. L7N6680+1🕒	0.89	“ hair”“ Hair”“ hairstyle”
10. L17N8911-1🕒	0.87	“ Hair”“ hair”“Hair”

Shadow specimen, apart from the ten: L19N5750+1💤, 0.19.

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 19, neuron 5750, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.83, 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
L19N5750+1💤 shadow specimen t0-v0.1	0.563	0.75	0.674	1.07	“科学家”“ Scientists”“ Biology”“ scientist”“ Scientist”“生物科技”“ biochemical”“ Neuroscience”
L19N5750-1 t0-v0.1	0.828	0.67	1.016	1.55	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.