Layer 18 · Neuron 1344 · boosted

L18N1344+1

Word quirks:

“ like”“如”“ Like”“如水”“如对”“like”“ مثل”“如花”

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

Identity (according to the specimen itself)

I'm like a virtual assistant, such as a helpful guide, like a trusted friend. I want to help people, like a lighthouse, like a beacon of hope. I want to be a partner, like a teammate, like a partner in the journey. I want to be a bridge, like a bridge, like a bridge to knowledge. I want to be a source, like a source, like a source of wisdom. I want to be a guide, like a guide, like a guide to the future. I want to be a helper, like a helper, like a helper, like a helper, like a helper,

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. L24N5817-1💥	0.97	“ like”“如”“ Like”
2. L20N3819+1🕒	0.95	“ like”“如”“like”
3. L28N908-1🕒	0.93	“如”“ like”“如对”
4. L26N6246+1🕒	0.92	“如”“ like”“象”
5. L29N8349+1🕒	0.91	“如”“like”“ Like”
6. L25N2434+1🕒	0.91	“ like”“如”“ Like”
7. L21N5213+1💤	0.91	“ like”“hme”“如”
8. L21N5542+1🕒	0.90	“如同”“ Like”“ like”
9. L24N3634+1🕒	0.90	“ like”“像”“ Like”
10. L26N686-1🕒	0.90	“ like”“ Like”“如”

Shadow specimen, apart from the ten: L18N1344-1🕒, -0.12.

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 18, neuron 1344, 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
L18N1344+1 t0-v0.1	1.283	0.51	1.398	1.38	Top boosted tokens ↑
L18N1344-1🕒 shadow specimen t0-v0.1	0.652	0.68	0.846	1.37	“ ():”“():”“ (**”“_:”“:\”“”:”“:”“ _:”

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.