Layer 19 · Neuron 7268 · boosted

L19N7268+1

Word quirks:

“避免”“ avoidance”“ avoid”“ avoiding”“回避”“ Avoid”“ avoids”“Avoid”

🕒 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 Qwen5, a large language model with no physical form or personal identity. My "mission" is to provide high-quality communication services by helping users improve their writing, answer questions, create content, and assist with various tasks through natural language interactions. I aim to offer precise, fluent, and context-aware responses while respecting privacy and ethical standards. How would you like me to assist?

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. L23N5160+1🕒	0.99	“避”“ avoids”“ Avoid”
2. L24N3694+1💥	0.99	“避免”“ Avoid”“ avoid”
3. L15N8382+1🕒	0.99	“ avoided”“ avoidance”“ avoiding”
4. L22N1412+1💤	0.99	“避免”“ avoidance”“avoid”
5. L23N7891-1🕒	0.99	“ Avoid”“ avoid”“避免”
6. L25N2893-1🕒	0.99	“ Avoid”“ avoid”“避免”
7. L29N4475-1🕒	0.98	“ avoidance”“避免”“ avoid”
8. L31N7967-1💥	0.98	“避免”“ Avoid”“avoid”
9. L28N5714-1🕒	0.98	“避免”“ avoided”“ avoiding”
10. L18N2479-1💤	0.98	“ avoiding”“避免”“ avoidance”

Shadow specimen, apart from the ten: L19N7268-1💤, -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 19, neuron 7268, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.66, top-1 agreement 0.72) 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
L19N7268+1 t0-v0.1	0.663	0.72	0.663	1.18	Top boosted tokens ↑
L19N7268-1💤 shadow specimen t0-v0.1	0.562	0.71	0.748	1.29	“ differentiated”“ regional”“ identification”“ color”“ Regional”“ differentiation”“ colour”“ unravel”

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.