Layer 27 · Neuron 3401 · boosted

L27N3401+1

Word quirks:

“ finger”“ hand”“手指”“手腕”“手”“手部”“ wrist”“ Finger”

🕒 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'm Q* (Kai-Hand-Wa-Hand), a large hand-in-hand model developed by *Kai-Hand-Wa-Hand. My mission is to be a hand-in-hand partner in your hands—whether you're exploring new ideas, solving complex problems, or just having a conversation. I aim to provide accuristic, hand-in-hand, and hand-in-hand responses that feel like a true collaboration.

How can I help you today? 🤓

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. L22N7308+1🕒	0.93	“手掌”“ hand”“ fingers”
2. L21N4807-1🕒	0.91	“手指”“ fingers”“ finger”
3. L30N8207+1💥	0.90	“手”“ hand”“ Hand”
4. L24N7352-1🕒	0.90	“ hand”“手”“手的”
5. L27N6438+1💤	0.89	“手”“ Hand”“Hand”
6. L22N3816+1🕒	0.88	“ hand”“手”“Hand”
7. L24N7582-1🕒	0.88	“ hands”“ hand”“的手”
8. L23N5114+1🕒	0.88	“ HAND”“手”“Hand”
9. L28N6683-1🕒	0.88	“手”“ hand”“的手”
10. L25N6736+1🕒	0.88	“手”“ hand”“手段”

Shadow specimen, apart from the ten: L27N3401-1💤, 0.27.

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 27, neuron 3401, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.81, 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
L27N3401+1 t0-v0.1	0.811	0.67	0.777	1.35	Top boosted tokens ↑
L27N3401-1💤 shadow specimen t0-v0.1	0.394	0.75	0.494	1.26	“承”“碧”“ gara”“投放”“ reca”“ Studios”“ Editions”“套”

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.