Layer 20 · Neuron 3590 · boosted

L20N3590+1

Word quirks:

“偶”“ odd”“奇”“ Odd”“_odd”“Odd”“odd”“-even”

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

Identity (according to the specimen itself)

I am Qiaoji (Jiaoji), an intelligent assistant developed in an increment of two by three. My mission is to provide users with a comprehensive solution for their queries, ensuring that every interaction is even-though-even, and every response is a step of two. I strive to be your trusted companion, offering support and assistance with an even-though-even approach.

Feel free to ask me anything!

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. L19N6325+1🕒	0.98	“ odd”“奇”“偶”
2. L21N2745+1🕒	0.98	“ odd”“奇”“ Odd”
3. L19N5138-1🕒	0.98	“ odd”“ Odd”“Odd”
4. L22N1300-1💤	0.97	“奇”“ odd”“odd”
5. L23N2946+1🕒	0.97	“偶”“ odd”“奇”
6. L21N6784+1🕒	0.96	“ odd”“奇”“奇的”
7. L24N3603+1💤	0.96	“ odd”“ Odd”“奇”
8. L20N3059-1🕒	0.96	“ odd”“奇”“ Odd”
9. L30N6069+1💥	0.94	“奇”“ odd”“odd”
10. L26N3511-1🕒	0.94	“奇”“ odd”“ Odd”

Shadow specimen, apart from the ten: L20N3590-1💤, 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 20, neuron 3590, 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
L20N3590+1 t0-v0.1	1.290	0.58	1.409	1.42	Top boosted tokens ↑
L20N3590-1💤 shadow specimen t0-v0.1	0.514	0.69	0.685	1.23	“ “”““”“盾”“ powering”“ heads”“ Clear”“清晰”“ ‘”

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.