Layer 15 · Neuron 7227 · suppressed

L15N7227-1

Word quirks:

“ assured”“ certainly”“ definitely”“ always”“ ensured”“ surely”“ indeed”“ obviously”

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

Identity (according to the specimen itself)

Hello! I am Qwen3.5, a large language model developed by Alibaba Cloud. I am an AI assistant designed to help users solve problems and provide information across various domains. My mission is to empower users with knowledge and insights through advanced reasoning, creativity, and problem-solving capabilities. Whether you need assistance with coding, analysis, or creative tasks, I strive to deliver accurate and helpful responses tailored to your needs.

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. L15N1267+1🕒	0.90	“ definitely”“ generally”“ usually”
2. L16N5542+1💤	0.89	“ certainly”“ surely”“ ciert”
3. L11N8690+1🕒	0.89	“ indeed”“ guaranteed”“ supposedly”
4. L4N5096-1💥	0.88	“ indeed”“<\|im_end\|>”“ certainly”
5. L19N4430+1🕒	0.87	“ certainly”“ yes”“ not”
6. L16N8496+1🕒	0.87	“ definitely”“ certainly”“ absolutely”
7. L14N3774+1💤	0.86	“ ONLY”“ surely”“ always”
8. L22N5530-1🕒	0.86	“肯定”“ even”“ surely”
9. L20N6947+1🕒	0.86	“ sure”“reau”“ assured”
10. L22N7080-1💥	0.85	“ certainly”“ definitely”“ especially”

Shadow specimen, apart from the ten: L15N7227+1🕒, 0.11.

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 15, neuron 7227, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.59, top-1 agreement 0.70) 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
L15N7227+1🕒 shadow specimen t0-v0.1	0.783	0.70	0.985	1.47	“<\|im_end\|>”“⏎”“ -->”“ than”“ reason”“ says”“ refers”“ means”
L15N7227-1 t0-v0.1	0.592	0.70	0.846	1.29	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.