Layer 10 · Neuron 6427 · suppressed

L10N6427-1

Word quirks:

“ **"”“ **”“ "**”“ **“”“ seems”“ basically”“ somewhat”“ description”

💤 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 Qwen, a large language model developed by Tongyi Lab. I am an AI system designed to assist with complex reasoning tasks, such as solving math problems, logical deductions, or generating creative content. My training data includes a vast array of information, and I can perform calculations or generate text based on user prompts.

My Mission:
To provide high-quality answers to questions that require specialized knowledge (like solving equations, writing stories, or coding) by breaking down a problem into smaller steps. I do not need to "know" everything myself; rather, my role is to act as the intelligent agent that

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. L10N2119-1🕒	0.90	“ "”“ "”“ basically”
2. L9N3492+1💤	0.84	“ “”“ description”“ ‘”
3. L10N1210+1💤	0.83	“ **"”“ usually”“ actually”
4. L12N3335+1💤	0.82	“ "”“ "”“ somehow”
5. L14N4562+1🕒	0.82	“ \"”“ „”“ “”
6. L9N5789-1🕒	0.82	“ "”“ "”“ ____”
7. L9N8075+1💤	0.82	“ basically”“ "”“ “”
8. L15N8566-1🕒	0.82	“ “”“ "”“ **”
9. L12N4296+1💤	0.82	“ "”“ “”“ **"”
10. L12N1805+1💤	0.82	“ ”“ related”“ "”

Shadow specimen, apart from the ten: L10N6427+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 10, neuron 6427, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.92, top-1 agreement 0.65) 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
L10N6427+1🕒 shadow specimen t0-v0.1	1.398	0.57	1.669	1.34	“ after”“ cleanup”“ afterwards”“ AFTER”“ afterward”“ final”“ completion”“ leftover”
L10N6427-1 t0-v0.1	0.923	0.65	1.199	1.32	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.