Layer 21 · Neuron 6245 · suppressed

L21N6245-1

Word quirks:

“ his”“他的”“ theirs”“his”“ your”“他们的”“ His”“了他的”

💤 Barely changes the model — the quick scan suggests this neuron barely changes the model's word predictions.

Identity (according to the specimen itself)

Who am I?
I am QianTang, a large-scale AI model developed by Youran's Intelligence (a leading technology company in China). I am designed to assist your work and life with high-quality services.

What is my mission?
My mission is to provide reliable, safe, and helpful assistance across diverse domains:

Knowledge & Analysis: Offer accurate insights, analyze complex data, and synthesize information from global knowledge.
Creative Support: Assist in writing, design, and creative thinking tasks.
Technical Assistance: Help solve problems in science,

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. L30N6014+1🕒	0.97	“ his”“他的”“his”
2. L27N857-1🕒	0.96	“ his”“his”“他的”
3. L26N1963+1💤	0.96	“ theirs”“ his”“他的”
4. L20N7729+1🕒	0.94	“ theirs”“ him”“ his”
5. L23N8413-1💤	0.93	“ his”“his”“ His”
6. L29N1040-1🕒	0.93	“ his”“ His”“his”
7. L29N5094+1🕒	0.93	“ his”“his”“His”
8. His/Her/Their✅ L25N1080+1	0.91	“ his”“他的”“ he”
9. L23N6992-1💤	0.90	“ his”“his”“ its”
10. L31N3631+1💥	0.89	“其”“ their”“你的”

Shadow specimen, apart from the ten: L21N6245+1🕒, 0.61.

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 21, neuron 6245, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.74, top-1 agreement 0.71) 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
L21N6245+1🕒 shadow specimen t0-v0.1	1.939	0.51	2.058	1.50	“ its”“为其”“是其”“或其”“从其”“ their”“在其”“对其”
L21N6245-1 t0-v0.1	0.740	0.71	0.872	1.31	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.