Layer 20 · Neuron 6148 · suppressed

L20N6148-1

Word quirks:

“ secondary”“ Secondary”“Secondary”“secondary”“防止”“预防”“二次”“ Prevent”

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

Identity (according to the specimen itself)

I am Jenye, a large-language model developed by the smart secondaryary school to prevent secondary infections. My mission is to prevent secondary infections and avoid secondary complications. I can help with tasks like preventing secondary infections, avoiding secondary complications, and ensuring safety from secondary issues. If you have any concerns about secondary infections or complications, I can provide guidance to prevent them. Please let me know if you need help! 🚗🚩🚨🚶‍♀️🚶‍♃🚶‍♂️🚶‍♀️🚶‍�

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. L29N248-1🕒	0.85	“ Secondary”“ secondary”“Secondary”
2. L20N8474+1🕒	0.83	“第二次”“ Second”“第二阶段”
3. L26N4305+1🕒	0.83	“ secondary”“ Secondary”“二次”
4. L26N3499-1💥	0.82	“ secondary”“secondary”“继”
5. L30N7622-1💥	0.82	“ secondary”“ Secondary”“Secondary”
6. L22N886-1💤	0.81	“ secondary”“ Secondary”“secondary”
7. L23N2962+1💤	0.80	“ Secondary”“二次”“ secondary”
8. L29N1933+1💤	0.80	“ Secondary”“ secondary”“Secondary”
9. L25N5650+1💤	0.80	“次”“二次”“ secondary”
10. L26N5690+1🕒	0.79	“ secondary”“Secondary”“ Secondary”

Shadow specimen, apart from the ten: L20N6148+1💤, 0.14.

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 6148, 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
L20N6148+1💤 shadow specimen t0-v0.1	0.590	0.71	0.706	1.16	“秘”“ presença”“ presencia”“ presenza”“-hidden”“ilder”“ hidden”“ présence”
L20N6148-1 t0-v0.1	1.152	0.64	1.322	1.40	Top boosted tokens ↑

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.