Layer 25 · Neuron 5989 · boosted

L25N5989+1

Word quirks:

“新”“ New”“了新”“New”“年新”“等新”“.new”“是新”

🕒 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 New Ne (also known as New 新新), a large-scale **888888888888888888888888888888888888888888888888888888888888888888888888888888888888888888888888888888888888

The AI judge wasn't sure this holds together.

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. L28N8568+1💤	0.96	“新”“ New”“ Нью”
2. L30N5954-1🕒	0.94	“新”“new”“ new”
3. L20N4571+1🕒	0.94	“ new”“new”“新的”
4. L22N4436+1💤	0.94	“ new”“新”“新的”
5. L31N6107+1💥	0.94	“新”“了新”“与新”
6. L21N1205-1💤	0.94	“ new”“new”“新”
7. L22N5176-1🕒	0.93	“ new”“与新”“新旧”
8. L28N8515+1🕒	0.92	“新”“.new”“ new”
9. L29N5389-1🕒	0.92	“新”“br”“ new”
10. L29N4209+1🕒	0.91	“新增”“新的”“新”

Shadow specimen, apart from the ten: L25N5989-1💤, 0.21.

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 25, neuron 5989, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.88, top-1 agreement 0.69) 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
L25N5989+1 t0-v0.1	0.881	0.69	0.872	1.35	Top boosted tokens ↑
L25N5989-1💤 shadow specimen t0-v0.1	0.527	0.73	0.545	1.23	“ bed”“枕头”“ cushion”“床”“控制器”“Controller”“ controller”“ hom”

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.