Layer 17 · Neuron 2243 · boosted

L17N2243+1

Word quirks:

“ called”“ known”“ named”“ termed”“ including”“ referred”“称作”“被称为”

🕒 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 an AI assistant developed by Alibaba, known as the "Q" series (specifically, I am Qwen, or "Tiankun" in Chinese). My name is Qwen (in English), and my Chinese name is Tiankun (in Chinese: 天库).

My mission is to assist users with a wide range of tasks, including answering questions, solving problems (such as logical reasoning and coding), creating content like writing and summarizing, and providing creative support. This includes helping with decision-making, such as planning tasks, analyzing data, and offering insights that can guide actions.

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. L16N1024+1🕒	0.95	“ referred”“ termed”“称作”
2. L19N3867+1🕒	0.94	“ termed”“ called”“称为”
3. L16N5387-1🕒	0.94	“ known”“ referred”“ called”
4. L17N4525-1🕒	0.92	“ called”“ known”“ termed”
5. L13N7040+1🕒	0.90	“ called”“ termed”“ known”
6. L19N920+1🕒	0.89	“ called”“ known”“ termed”
7. L20N1849-1🕒	0.89	“ called”“ Called”“called”
8. L12N7444+1🕒	0.88	“ called”“ known”“ referred”
9. L19N2357+1🕒	0.88	“ called”“ represented”“ labeled”
10. L11N6428-1🕒	0.88	“ called”“ termed”“ known”

Shadow specimen, apart from the ten: L17N2243-1💤, -0.00.

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 17, neuron 2243, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.80, top-1 agreement 0.67) 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
L17N2243+1 t0-v0.1	0.796	0.67	0.895	1.31	Top boosted tokens ↑
L17N2243-1💤 shadow specimen t0-v0.1	0.425	0.75	0.528	1.15	“ —”“——”“—”“—who”“—at”“—for”“—it”“ ——”

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.