Layer 3 · Neuron 6991 · suppressed

L3N6991-1

Word quirks:

“ ground”“ rests”“ onto”“ support”“ grounded”“ resting”“ repose”“ supports”

🕒 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 designed to provide support and guidance. My mission is to offer assistance and 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. L6N4006+1🕒	0.73	“ sleep”“ position”“ sleeping”
2. L4N7735-1🕒	0.72	“ equilibrium”“ repose”“ resting”
3. L19N5204-1🕒	0.71	“ zero”“ ground”“零”
4. L2N4868+1💥	0.68	“ repose”“ rest”“<\|im_end\|>”
5. L29N2879+1💥	0.67	“ resting”“ rests”“ rested”
6. L5N7745+1🕒	0.67	“ rest”“ rested”“ relaxation”
7. L8N8820-1🕒	0.66	“ setup”“ height”“ rested”
8. L20N7605-1🕒	0.66	“ rests”“ resting”“rest”
9. L21N3861-1💤	0.66	“ rests”“ resting”“-rest”
10. L23N3763-1🕒	0.66	“ rest”“Rest”“ Rest”

Shadow specimen, apart from the ten: L3N6991+1🕒, -0.26.

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 3, neuron 6991, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 2.19, top-1 agreement 0.42) 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
L3N6991+1🕒 shadow specimen t0-v0.1	1.305	0.51	1.563	1.81	“ -”“"!”“ !”“!"”“ --”“:"”“?\”“!”
L3N6991-1 t0-v0.1	2.195	0.42	2.601	2.18	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.