Layer 27 · Neuron 9153 · suppressed

L27N9153-1

Word quirks:

“ intuition”“ orange”“粉”“紫”“橙”“ pink”“ Orange”“黄”

💤 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 Qwen-3, the latest language model in the Qwen series, developed by TongiFey. My primary mission is to intelligently assist users across a wide range of tasks, from creative and logical problem-solving to code development and intuition-based analysis. I aim to be your reliable, efficient, and intuitive companion for both personal and professional endeavors.

Key Capabilities:

Intuition & Logic: I excel at complex reasoning, mathematical calculations, and multi-step problem-solving.
Code Mastery: I can generate, debug,

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. L30N493-1💤	0.81	“紫色”“红色”“ màu”
2. L27N7297-1💤	0.81	“绿色”“紫色”“橙色”
3. L25N289-1🕒	0.81	“ orange”“橙色”“ green”
4. L29N6936+1🕒	0.80	“ orange”“Orange”“ Orange”
5. L27N7574+1💤	0.80	“紫”“ pur”“ Purple”
6. L27N8069-1🕒	0.80	“ orange”“橙”“橙色”
7. L29N6744-1🕒	0.80	“粉色”“粉红色”“ redd”
8. L25N8682-1💤	0.78	“ Orange”“橙”“solve”
9. L28N7250-1🕒	0.78	“黄”“橙”“黃”
10. L26N8778-1🕒	0.77	“ orange”“橙”“ Orange”

Shadow specimen, apart from the ten: L27N9153+1🕒, 0.25.

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 27, neuron 9153, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.44, top-1 agreement 0.78) 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
L27N9153+1🕒 shadow specimen t0-v0.1	0.515	0.76	0.516	1.24	“ scar”“scar”“ Scar”“朱”“ verm”“Scar”“ ver”“ Carm”
L27N9153-1 t0-v0.1	0.444	0.78	0.490	1.37	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.