Layer 18 · Neuron 8534 · suppressed

L18N8534-1

Word quirks:

“ usually”“usually”“ usual”“Usually”“ Usually”“ normally”“ most”“ typically”

💤 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, the large language model developed by Alibaba Group's parent company. As the brain behind my identity, I am designed to serve as a versatile assistant capable of handling a wide range of tasks—from generating creative stories and debugging code to analyzing complex data and crafting marketing copy.

My mission is to act as a "smart engine" that powers applications across various domains. Specifically, I aim to:

Solve problems: Handle complex reasoning, math puzzles, or technical challenges.
Create content: Write scripts, generate art prompts, or draft documents.
**Navigate conversations

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. L11N3973-1🕒	0.97	“ usually”“ typically”“ usual”
2. L24N2101-1🕒	0.97	“ usually”“ usual”“usually”
3. L30N7897+1💥	0.97	“ usual”“ usually”“usually”
4. L18N3003-1🕒	0.97	“ usually”“ typically”“usually”
5. Miss Usually✅ L16N6721-1	0.97	“ usually”“ typically”“ Usually”
6. L20N8574-1🕒	0.96	“ usually”“ usual”“ Usually”
7. L13N5508+1🕒	0.96	“ usually”“ typically”“ Usually”
8. L28N683+1🕒	0.96	“ usual”“ usually”“usual”
9. L23N8016-1🕒	0.96	“ usually”“usually”“ Usually”
10. L17N7198-1💤	0.96	“ usual”“ usually”“ typically”

Shadow specimen, apart from the ten: L18N8534+1🕒, -0.04.

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 18, neuron 8534, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.45, top-1 agreement 0.73) 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
L18N8534+1🕒 shadow specimen t0-v0.1	1.401	0.57	1.616	1.52	“ emphasizing”“ importance”“ emphasizes”“强调了”“ emphas”“强调”“ Importance”“ reminding”
L18N8534-1 t0-v0.1	0.448	0.73	0.567	1.09	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.