Layer 7 · Neuron 6259 · boosted

L7N6259+1

Word quirks:

“ burial”“ cemetery”“墓”“ graves”“墓地”“ grave”“墓碑”“ tomb”

🕒 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 an AI assistant, and my mission is to provide a resting place for those who have passed away.

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. L5N7964+1🕒	0.97	“ cemetery”“ burial”“ graves”
2. L2N3290-1💥	0.95	“ burial”“ grave”“ graves”
3. The Gravekeeper✅ L6N6001-1	0.91	“ burial”“ buried”“ tomb”
4. L30N7943-1🕒	0.91	“葬”“墓”“ burial”
5. L19N2599+1🕒	0.87	“ burial”“ buried”“埋葬”
6. L8N4308+1🕒	0.83	“ burial”“ ancient”“葬”
7. L18N9071-1🕒	0.75	“ funeral”“ Funeral”“死神”
8. L9N4177+1🕒	0.72	“ burial”“ corpse”“遗体”
9. L18N7175-1🕒	0.72	“殉”“ memorial”“ martyr”
10. L18N8005-1🕒	0.71	“ deceased”“ tomb”“ ancestors”

Shadow specimen, apart from the ten: L7N6259-1🕒, -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 7, neuron 6259, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 1.65, top-1 agreement 0.49) 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
L7N6259+1 t0-v0.1	1.651	0.49	1.960	1.62	Top boosted tokens ↑
L7N6259-1🕒 shadow specimen t0-v0.1	1.101	0.61	1.325	1.48	“ confirmation”“ response”“ feedback”“ payment”“ confirming”“ request”“ update”“ customer”

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.