Layer 14 · Neuron 7270 · suppressed

L14N7270-1

Word quirks:

“ description”“ descriptions”“ described”“ describing”“ similar”“描述”“ Description”“ descri”

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

Identity (according to the specimen itself)

I am Qwen3.5, a large language model developed by Alibaba Cloud's Tongyi Qianwen series. My name is Tongyi Qianwen. If you need to refer to me, I can also use my developer name or other identifiers.

If you need to describe my features, I can help you with:

Knowledge Base: I have extensive knowledge of various topics, including books, authors, and historical events.
Mission: I am designed to assist users in completing tasks efficiently.
Capabilities: I support multi-modal interactions (e.g., images, documents) 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. L15N5305+1🕒	0.98	“ described”“ description”“ describes”
2. L8N1012+1💤	0.98	“ description”“ descriptions”“ describing”
3. L11N3051-1🕒	0.98	“ description”“ describing”“ descriptions”
4. L13N4916-1🕒	0.97	“ describes”“ description”“ describing”
5. L26N6618-1🕒	0.97	“ describe”“ describes”“ description”
6. L17N8003+1🕒	0.97	“ description”“ describing”“描述”
7. L12N2345+1💥	0.97	“ description”“ describe”“ Description”
8. L18N5709+1🕒	0.97	“ description”“ descriptions”“ described”
9. L17N7479+1💤	0.97	“ described”“ description”“ descriptions”
10. L26N3747+1🕒	0.96	“ Description”“Description”“ description”

Shadow specimen, apart from the ten: L14N7270+1🕒, 0.06.

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 14, neuron 7270, 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 0.66) 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
L14N7270+1🕒 shadow specimen t0-v0.1	1.072	0.59	1.362	1.58	“ actually”“ does”“ typically”“ provides”“?\”“ is”“ basically”“ doesn”
L14N7270-1 t0-v0.1	0.926	0.66	1.233	1.30	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.