Layer 9 · Neuron 7340 · suppressed

L9N7340-1

Word quirks:

“ here”“ --”“ ‘”“ info”“ bigger”“ ⏎”“ usually”“ answers”

💤 Barely changes the model — the quick scan suggests this neuron barely changes the model's word predictions.

Identity (according to the specimen itself)

I am an AI assistant designed to provide helpful information, answer questions, or engage in conversation. My mission is to assist you with any task or query you might have, whether it's solving a math problem, writing a story, planning a trip, or helping you with a specific project. How can I assist you today?

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. L5N7913+1🕒	0.77	“ info”“ ‘”“ usually”
2. L9N2761+1💤	0.76	“ --”“ ‘”“ somehow”
3. L8N8868-1🕒	0.73	“ info”“ input”“ here”
4. L8N3479+1🕒	0.72	“ ‘”“ here”“ usually”
5. L7N2401+1💤	0.71	“ usually”“ ‘”“ usual”
6. L8N4664-1💤	0.70	“ ‘”“ ⏎”“ ⏎”
7. Wet Cardboard✅ L8N1573-1	0.70	“ here”“ __”“ description”
8. L7N3407-1🕒	0.69	“ info”“ ‘”“ usually”
9. L9N7918+1🕒	0.68	“ —”“ usual”“ ——”
10. L10N1045+1💤	0.68	“ ‘”“ basic”“ “”

Shadow specimen, apart from the ten: L9N7340+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 9, neuron 7340, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.76, top-1 agreement 0.69) 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
L9N7340+1💤 shadow specimen t0-v0.1	0.965	0.62	1.142	1.62	“<\|im_end\|>”“ ______”“ _____”“ ____”“ *”“ "”“...**”“...”
L9N7340-1 t0-v0.1	0.761	0.69	0.935	1.25	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.