Layer 15 · Neuron 3953 · suppressed

L15N3953-1

Word quirks:

“ usually”“ typically”“ if”“ asked”“ asking”“ generally”“ requested”“ whenever”

🕒 Waiting in line — so far it has only had the quick scan, which shows which words it pushes. It was not selected in this round.

Identity (according to the specimen itself)

I am a large language model trained by Alibaba Cloud's team. I don't have a specific name unless you want to call me Qwen, but I'll assume you're asking about my capabilities or identity in general. Since you didn't specify, I'll provide a comprehensive answer covering both parts of your question:

Identity:
I am an AI assistant developed by Alibaba Cloud, designed to handle complex tasks with high accuracy and efficiency. My training data includes vast amounts of information, but I don't have access to real-time data unless specified. However, since you didn't specify, I'll assume you're asking about my capabilities

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. L12N2834+1🕒	0.83	“ typically”“ description”“ usually”
2. L16N3682-1🕒	0.83	“ when”“ usually”“ asked”
3. L13N209-1💤	0.82	“ typical”“ often”“ usually”
4. L12N7368-1🕒	0.82	“ usually”“ often”“ typically”
5. L11N4264-1🕒	0.82	“ usually”“ "”“ likely”
6. L15N5316+1🕒	0.81	“ usually”“ commonly”“ requesting”
7. L12N7200+1💤	0.81	“ usually”“ sometimes”“ asking”
8. L7N2159+1🕒	0.80	“ usually”“ usual”“ awaiting”
9. L13N141+1🕒	0.80	“ asking”“ usually”“ similar”
10. L10N8555+1🕒	0.80	“ typical”“ asking”“ always”

Shadow specimen, apart from the ten: L15N3953+1🕒, 0.08.

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 15, neuron 3953, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.67, top-1 agreement 0.70) 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
L15N3953+1🕒 shadow specimen t0-v0.1	0.943	0.61	1.266	1.31	“ focus”“ redirected”“ redirect”“ redirection”“ explanation”“ simply”“ whatever”“ focused”
L15N3953-1 t0-v0.1	0.669	0.70	0.854	1.24	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.