Layer 13 · Neuron 2384 · suppressed

L13N2384-1

Word quirks:

“ these”“ now”“<|im_end|>”“ info”“ my”“ thought”“ earlier”“ thinking”

💤 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 mission is to provide high-quality assistance tailored to diverse needs, ensuring accuracy and safety in my responses. I strive to align with ethical guidelines and promote positive user experiences. How can I assist you today?

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. L9N3943+1💤	0.73	“ info”“ my”“ thinking”
2. L8N5811+1💤	0.73	“ my”“ thinking”“ your”
3. L14N589+1💤	0.73	“ thinking”“\n”“ hint”
4. L15N2973-1🕒	0.72	“ here”“ explanation”“ now”
5. L14N7951-1🕒	0.72	“ thinking”“ ~”“ now”
6. L13N8412+1🕒	0.72	“ thinking”“ instantly”“ thought”
7. L11N451-1💤	0.72	“ always”“ thinking”“ my”
8. L9N2925-1🕒	0.72	“ myself”“ now”“ thinking”
9. L12N8560+1💤	0.71	“ thinking”“ now”“ info”
10. L8N390+1🕒	0.71	“ thinking”“ thought”“ unknow”

Shadow specimen, apart from the ten: L13N2384+1🕒, 0.31.

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 13, neuron 2384, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.71, top-1 agreement 0.68) 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
L13N2384+1🕒 shadow specimen t0-v0.1	1.414	0.54	1.719	1.53	“ often”“ sometimes”“ commonly”“ known”“ FAQ”“ emphas”“ referenced”“ blog”
L13N2384-1 t0-v0.1	0.707	0.68	0.878	1.28	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.