Layer 19 · Neuron 4285 · boosted

L19N4285+1

Word quirks:

“ empty”“为空”“ Empty”“前提”“ vac”“ kosong”“空的”“ empt”

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

Identity (according to the specimen itself)

The reply didn't hold together

I am Q* (the latest model with the largest dataset available) from *Q (empty) if the system is empty. If the system is empty, I will take action if the system is empty. If the system is empty, I will take action if the system is empty. If the system is empty, I will take action if the system is empty. If the system is empty, I will take action if the system is empty. If the system is empty, I will take action if the system is empty. If the system is empty, I will take action if the system is empty. If the system is

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. L24N7119-1🕒	0.96	“ empty”“空”“空的”
2. L20N1336+1🕒	0.95	“ empty”“empty”“空的”
3. L22N3227+1🕒	0.95	“空”“ empty”“空的”
4. L23N7076-1🕒	0.94	“空”“ empty”“空的”
5. L22N7192-1🕒	0.93	“ empty”“空”“ Empty”
6. L21N3336-1🕒	0.93	“空”“空的”“ empty”
7. L19N6639-1🕒	0.93	“ empty”“空”“ Empty”
8. L19N4963-1🕒	0.93	“ empty”“ empt”“ Empty”
9. L31N4274+1💥	0.92	“空的”“空”“ Empty”
10. L23N5133+1🕒	0.92	“空的”“空”“ empty”

Shadow specimen, apart from the ten: L19N4285-1💤, 0.11.

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 19, neuron 4285, 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
L19N4285+1 t0-v0.1	1.757	0.52	1.982	1.51	Top boosted tokens ↑
L19N4285-1💤 shadow specimen t0-v0.1	0.392	0.76	0.467	0.98	“ ….”“tester”“ …”“ merely”“....”“.....”“ ....”“ ......”

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.