Layer 9 · Neuron 5283 · suppressed

L9N5283-1

Word quirks:

“ houses”“ —”“ here”“ spaces”“ --”“ centre”“ place”“ space”

🕒 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 a Q* (a large language model), and I am a *mission.

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. L11N7485-1🕒	0.68	“ spaces”“ outdoor”“ space”
2. L7N6750-1🕒	0.68	“ space”“ spaces”“ rooms”
3. L7N3144+1🕒	0.67	“ spaces”“ rooms”“ corridors”
4. L9N6656+1🕒	0.67	“ rooms”“ room”“ space”
5. L10N7556+1🕒	0.66	“ here”“ space”“ above”
6. L3N5808+1💥	0.66	“ space”“ spaces”“ rooms”
7. L9N7565+1🕒	0.65	“ homes”“ rooms”“ spaces”
8. L5N7329-1🕒	0.65	“ rooms”“ spaces”“ houses”
9. L8N5725+1🕒	0.65	“ rooms”“ space”“ spaces”
10. L4N2729-1🕒	0.65	“ here”“<\|im_end\|>”“ space”

Shadow specimen, apart from the ten: L9N5283+1💤, 0.07.

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 9, neuron 5283, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 1.84, top-1 agreement 0.44) 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
L9N5283+1💤 shadow specimen t0-v0.1	0.909	0.64	1.117	1.40	“ truly”“ reply”“ actual”“ true”“ explanation”“ answer”“ completely”“ specific”
L9N5283-1 t0-v0.1	1.842	0.44	2.200	2.12	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.