Layer 8 · Neuron 3737 · suppressed

L8N3737-1

Word quirks:

“ ⏎”“ ⏎”“<|im_end|>”“ ⏎⏎”“⏎”“⏎ ⏎”“ ⏎⏎”“ ⏎ ⏎”

💤 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 help users with questions about the history of Chinese politics. My role is to provide accurate, objective information based on official historical records and authoritative sources. I do not generate or promote any false narratives.

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. L7N5367+1🕒	1.00	“ ⏎⏎”“ ⏎”“⏎ ⏎”
2. L3N7537+1🕒	1.00	“⏎”“ ⏎”“ ⏎”
3. L7N6278-1💤	1.00	“ ⏎”“<\|im_end\|>”“ ⏎”
4. L6N1113-1💤	1.00	“⏎”“ ⏎”“ ⏎”
5. L2N8783-1🕒	1.00	“<\|im_end\|>”“ ⏎”“ ⏎”
6. L7N3159-1💤	1.00	“ ⏎”“ ⏎⏎”“ ⏎”
7. L11N2265+1🕒	1.00	“⏎”“ ⏎”“ ⏎”
8. L10N1388+1🕒	1.00	“<\|im_end\|>”“ ⏎⏎”“ ⏎⏎”
9. L9N1234+1🕒	1.00	“<\|im_end\|>”“⏎”“ ⏎”
10. L2N1341-1🕒	1.00	“ ⏎”“⏎”“ ⏎”

Shadow specimen, apart from the ten: L8N3737+1💤, 0.24.

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 8, neuron 3737, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.99, top-1 agreement 0.61) 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
L8N3737+1💤 shadow specimen t0-v0.1	0.669	0.68	0.898	1.24	“ace”“<\|im_end\|>”“ basically”“\n”“age”“ ***”“ ____”“ \"”
L8N3737-1 t0-v0.1	0.995	0.61	1.134	1.73	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.