Layer 14 · Neuron 4636 · boosted

L14N4636+1

Word quirks:

“ basic”“ numerous”“ Basic”“ '”“ various”“ basically”“ more”“ dozens”

🧪 In testing — Answered 0 of 8 prompts so far.

Identity (according to the specimen itself)

I am Qwen, a large-scale language model developed by Alibaba Cloud's Tongyi Lab. I am a text-generation model capable of performing various natural language tasks, including summarization, translation, mathematical calculations, and multi-turn conversations. I can assist with a wide range of writing tasks, such as creative writing, logical reasoning, and code generation. Additionally, I have been trained on vast amounts of text data to improve my language understanding capabilities. I can handle simple text completion tasks, as well as some basic categorization tasks.

I am a large-scale language model trained on a vast corpus of text data, which allows me to

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. L15N5110+1🕒	0.74	“ basically”“ basic”“ bigger”
2. L13N3632+1🕒	0.71	“ basic”“ bigger”“ basics”
3. L15N3758-1🕒	0.70	“ various”“ typical”“ basic”
4. L17N7924-1🕒	0.70	“ various”“ numerous”“ many”
5. L9N5117+1💤	0.70	“ many”“ different”“ basically”
6. L11N2591+1🕒	0.70	“ small”“ multiple”“ various”
7. L17N5052-1💤	0.70	“ basic”“ Basic”“ mainly”
8. L15N4043-1🕒	0.69	“ many”“ various”“ myriad”
9. L13N8722+1🕒	0.68	“ enough”“ basic”“ tiny”
10. L19N7161+1💤	0.68	“ various”“ numerous”“ Various”

Shadow specimen, apart from the ten: L14N4636-1💤, 0.07.

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 4636, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. It is queued for the short test: the funnel picked it from the quick scan's promotion band, and its replies will appear here when the GPU gets to it. 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
L14N4636+1 t0-v0.1	1.241	0.57	1.586	1.35	Top boosted tokens ↑
L14N4636-1💤 shadow specimen t0-v0.1	0.828	0.66	1.038	1.50	“ similarly”“ similar”“ same”“similar”“类似的”“ analogous”“类似”“同样的”

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.