Layer 14 · Neuron 4564 · boosted

L14N4564+1

Word quirks:

“ answers”“ questions”“ answer”“ answered”“ question”“ answering”“ asking”“ Answers”

💤 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. My primary goal is to provide comprehensive answers to questions across various domains, such as:

What is the best way to create content?
"How do I write a story?"
"How do I write a poem?"
"How do I write a story?"
"How do I write a story?"
"How do I write a story?"
"How do I write a story?"
"How do I write a story?"
"How

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. L13N2177-1🕒	0.98	“ answer”“ answers”“ question”
2. L8N424-1💤	0.98	“ questions”“ question”“ answers”
3. L17N5804-1🕒	0.98	“ answering”“ answer”“ answers”
4. L11N6086+1🕒	0.98	“ questions”“ question”“ answered”
5. L8N2271-1🕒	0.97	“ answer”“ question”“ answers”
6. Pop Quiz✅ L8N1567+1	0.96	“ answer”“ question”“ answers”
7. L12N8102-1🕒	0.96	“ question”“ answer”“ questions”
8. L12N2583-1💤	0.96	“ question”“ answer”“ questions”
9. L15N636-1🕒	0.96	“ asking”“ answer”“ answers”
10. L4N3882-1🕒	0.95	“ question”“ questions”“ reply”

Shadow specimen, apart from the ten: L14N4564-1🕒, 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 4564, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.94, top-1 agreement 0.59) 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
L14N4564+1 t0-v0.1	0.940	0.59	1.207	1.49	Top boosted tokens ↑
L14N4564-1🕒 shadow specimen t0-v0.1	1.414	0.49	1.717	1.54	“ emphasizing”“ emphasis”“ ensuring”“ representing”“ emphasizes”“ protecting”“ emphas”“ ensures”

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.